Drinking Water Infrastructure Needs Survey and Assessment
Sixth Report to Congress
Office of Water (4606M)
EPA 816-K-17-002
March 2018
Drinking Water Infrastructure Needs Survey and Assessment
Sixth Report to Congress
U.S. Environmental Protection Agency
Office of Water
Office of Ground Water and Drinking Water
Drinking Water Protection Division
Washington, D.C. 20460
Contents
Drinking Water Infrastructure Needs Survey and Assessment ............................................................................... 1
Sixth Report to Congress .................................................................................................................................... 1
Executive Summary ................................................................................................................................................ 9
Total National Need ............................................................................................................................................ 9
National Need Compared to Previous Needs Assessments .............................................................................. 10
Individual State Need ........................................................................................................................................ 10
Regulatory Need ............................................................................................................................................... 11
Small System Need ........................................................................................................................................... 11
Needs of American Indian and Alaska Native Village Water Systems ............................................................ 11
Water Industry Capital Investment Planning and Documentation of Needs .................................................... 12
Chapter 1: Findings - National Need .................................................................................................................... 12
2015 Total National Need ................................................................................................................................. 13
2015 Total National Need Compared to EPA’s Previous Assessments ........................................................... 15
Total National Need by Project Category ......................................................................................................... 20
New Infrastructure Needs versus Needs Associated with Existing Infrastructure ........................................... 28
Need by System Size ........................................................................................................................................ 29
Needs Associated with SDWA Regulations ..................................................................................................... 30
Chapter 2: Findings - State Need .......................................................................................................................... 35
State-Specific Needs ......................................................................................................................................... 35
Needs of Water Systems in U.S. Territories ..................................................................................................... 39
Changes in State-Specific Need through Assessment Cycles ........................................................................... 40
Chapter 3: American Indian and Alaska Native Village Need ............................................................................. 42
American Indian and Alaska Native Village-Specific Needs ........................................................................... 43
American Indian Needs..................................................................................................................................... 44
Alaska Native Village Needs ............................................................................................................................ 44
Chapter 4 Conclusions .......................................................................................................................................... 46
Appendix A - Survey Methods ............................................................................................................................. 47
Assessing the Needs of Water Systems in States and U.S. Territories ............................................................. 47
Assessing the Need for Small Systems ............................................................................................................. 54
Assessing the Need of Not-for-Profit Noncommunity Water Systems ............................................................ 54
Assessing the Need of American Indian and Alaska Native Village Water Systems ....................................... 55
Appendix B - Data Collection ............................................................................................................................... 56
Survey Instrument ............................................................................................................................................. 57
Project Documentation ...................................................................................................................................... 57
Cost Estimates and Modeling ........................................................................................................................... 58
Website and Database ....................................................................................................................................... 58
Quality Assurance ............................................................................................................................................. 59
Continuing Evolution of the DWINSA ............................................................................................................. 60
Appendix C Statutory and Regulatory Criteria and Policies ............................................................................. 62
Project Allowability .......................................................................................................................................... 62
Documentation Requirements ........................................................................................................................... 63
Appendix D - Accuracy, Precision, and Uncertainty ............................................................................................ 68
Uncertainty ........................................................................................................................................................ 68
Precision ............................................................................................................................................................ 68
Bias ................................................................................................................................................................... 69
Appendix E - Summary of Findings for State-Systems Serving 10,000 and Fewer People ................................. 71
Community Water Systems Serving 10,000 and Fewer People ....................................................................... 71
Glossary ................................................................................................................................................................ 73
Acknowledgments
Many dedicated individuals contributed to the 2015 Drinking Water Infrastructure Needs Survey and Assessment.
We would like to thank the states for their active participation and continuing interest in the project. And most
importantly, we would like to thank the operators and managers of the thousands of water systems who spent
their valuable time providing information for the Assessment.
9
Executive Summary
Total National Need
The U.S. Environmental Protection Agency’s (EPA’s) sixth national
assessment of public water system infrastructure needs shows a total
20-year capital improvement need of $472.6 billion. This estimate
represents Drinking Water State Revolving Fund (DWSRF) - eligible
infrastructure projects necessary from January 1, 2015, through
December 31, 2034, for water systems to continue to provide safe
drinking water to the public. The national total comprises the
infrastructure investment needs of the nation’s approximately 49,250
community water systems (CWSs),
1
21,400 not-for-profit
noncommunity water systems (NPNCWSs), American Indian water
systems and Alaska Native Village water systems.
2
The findings are
based on the 2015 Drinking Water Infrastructure Needs Survey and
Assessment (DWINSA or Assessment), which relied primarily on a
statistical survey of public water systems. The survey response rate was
99.7 percent (2,592 responses from 2,600 systems surveyed), the highest response rate in the history of the
Assessment, providing a high degree of confidence in the statistical precision of the Assessment’s findings.
The estimate covers infrastructure needs that are eligible for (but
not necessarily financed by) the DWSRF, including the
installation of new drinking water infrastructure and the
rehabilitation, expansion, or replacement of existing
infrastructure. The results of the Assessment are used to allocate
the DWSRF capitalization grants to the states for their DWSRF
programs and to the EPA regional offices for American Indian and
Alaska Native Village systems. The reported projects may be
needed to address existing infrastructure that is deteriorated or
undersized, ensure compliance with regulations, provide system
resilience, improve energy efficiency, or improve cost
effectiveness. Cost estimates reflect comprehensive construction
costs including engineering and design, purchase of raw materials
and equipment, construction and installation labor, and final
inspection.
1
The estimated 49,250 CWSs was derived from the December 2013 Safe Drinking Water Information System (SDWIS) freeze for
active community water systems, excluding federally-owned systems.
2
The inventory of American Indian and Alaska Native Village water systems was derived from the December 2009 SDWIS freeze.
Authority, Purpose, and History
The 1996 Safe Drinking Water Act
Amendments mandated that EPA conduct
an assessment of the nation’s public water
systems’ infrastructure needs every four
years and use the findings to allocate
DWSRF capitalization grants to states.
The DWSRF was established to help
public water systems obtain financing for
improvements necessary to protect public
health and comply with drinking water
regulations. From 1997 to 2016, states
provided $32.5 billion through DWSRF
programs to water systems for 13,183
projects.
$472.6 Billion is Needed
The nation’s drinking water utilities
need $472.6 billion in infrastructure
investments over the next 20 years
for thousands of miles of pipe as
well as thousands of treatment
plants, storage tanks, and other key
assets to ensure the public health,
security, and economic well-being
of our cities, towns, and
communities.
10
EPA recognizes that there are significant water system needs that are generally ineligible for DWSRF funding,
such as raw water dams and reservoirs, projects related primarily to population growth, and water system
operation and maintenance costs. Because the Assessment is directly associated with the allocation of DWSRF
capitalization grants to states and tribal set-aside funds to EPA regional offices, needs ineligible for DWSRF
funding are not included in the estimate.
National Need Compared to Previous Needs Assessments
EPA conducted five previous Assessments in 1995, 1999, 2003,
2007, and 2011. Exhibit ES.1 presents the total national need for
the Assessments, adjusted to 2015 dollars. The 67 percent increase
in need between 1999 and 2003 was the result of a focus on better
capturing the full 20-year need of surveyed systems by including
an asset inventory-based approach to identify longer-term
infrastructure replacement and rehabilitation needs. Beginning
with the 2003 Assessment, a water system’s inventory of existing
infrastructure assets, along with a simple statement of need, would
be considered as survey-generated documentation sufficient to
justify certain replacement and rehabilitation needs (as described
in Appendix C). The asset inventory-based approach continued in subsequent Assessments with data generally
demonstrating increases in inventoried assets on a national basis. Total national need increased by about 1 percent
in both the 2007 and 2011 Assessments, essentially the same statistical result as the 2003 findings. The 2015
Assessment, however, reveals a 10 percent increase in the estimate of total national need with survey data
indicating the largest increase in rehabilitation and replacement needs for existing infrastructure, specifically in
the water transmission and distribution project category. This increase was seen in both medium and large sized
systems.
For the 2015 Assessment, EPA adjusted its statistical methodology slightly to make use of a modified panel
approach for medium-sized systems, in which EPA resurveyed a large portion of the systems that participated in
the 2011 Assessment (as described in Appendix A). This approach facilitated the process of participation for the
medium sized systems by allowing a majority of these systems to update their responses from the 2011
Assessment, which led to more complete inventories of needs in the 2015 Assessment than in prior previous
Assessments.
Exhibit ES.1: DWINSA Comparison of 20-year National Need (in billions of January 2015 dollars)
Year
1995
2003
2007
2011
2015
National Need
$253.6
$419.4
$423.7
$428.6
$472.6
Individual State Need
As presented in Chapter 2, the 2015 Assessment shows significant changes in some states’ needs from the
previous Assessment. These changes will result in modifications to individual states’ DWSRF allotments.
Throughout the history of the DWINSA, each state’s needs change from one survey to the next, sometimes
significantly. Such changes in state need can be attributed to a number of factors, including expected changes in
the status of projects (i.e., recently planned projects versus initiated or completed projects).
Fully Capturing Longer-Term Needs
The 2015 Assessment continued the
practice used since the 2003 effort to
better capture 20-year investment needs
by including an asset inventory-based
approach to identify long-term
infrastructure replacement and
rehabilitation needs.
11
Regulatory Need
The findings of the 2015 Assessment indicate that the need
associated directly with Safe Drinking Water Act (SDWA)
regulatory compliance remains a relatively small percentage, just
over 12 percent of the total national need. Most water system
needs are not directly related to violations of, or compliance with,
SDWA regulations. Most needs, such as the replacement or
rehabilitation of leaking water mains, are ongoing investments that
systems must make to continue delivering safe drinking water to
their customers.
Small System Need
The 2015 Assessment includes a total national need of $74.4
billion for small systems. Small systems are defined as serving
3,300 or fewer people. For the 2015 Assessment, EPA did not
directly survey small systems but estimated the infrastructure
investment needs for these systems by adjusting the findings from
the field survey completed for small systems in states, Puerto
Rico, and the U.S. territories for the 2007 Assessment. In making the adjustment, EPA applied 2015 cost models
using the current inventory of small systems.
Needs of American Indian and Alaska Native Village Water Systems
As presented in Chapter 3, the needs of water systems serving American Indians and Alaska Native Villages total
$3.8 billion. For the 2015 Assessment, EPA did not directly survey these systems but estimated the infrastructure
investment needs by adjusting the findings from the 2011 Assessment. In making the adjustment, EPA applied
2015 cost models using the 2011 inventory of systems. This need represents a small percentage of the nation’s
total drinking water infrastructure need. This need is, however, associated with higher average per-household
costs due to unique challenges that many of these water systems face. These public water systems are almost all
small and are often located in remote rural areas, some in areas with permafrost, and the communities served may
have households that lack access to the public water supply. These conditions present special challenges for
providing drinking water service.
Beyond providing the basis for allotting the Drinking Water State Revolving Fund, the Assessment and Report
to Congress have been instrumental in helping systems to more fully inventory their assets and to look out over
a long-term planning horizon. The evolution in water system capital planning continues as systems increasingly
adopt asset management strategies that will enable them to most cost-effectively ensure delivery of service that
protects public health and supports economic prosperity.” Peter Grevatt, Director, EPA Office of Ground
Water and Drinking Water.
Data Presentation
Throughout this report, information
presented in the tables and figures is
derived from DWINSA survey data from
the referenced year unless otherwise
noted. Data for small systems (serving
3,300 or fewer people) is extrapolated
from information collected in 2007; data
for American Indian and Alaska Native
Village systems is extrapolated from
information collected in 2011; and data
for Not-for-Profit Noncommunity water
systems is extrapolated from information
collected in 1999.
12
Water Industry Capital Investment Planning and Documentation of Needs
Systems submitted a variety of planning documents and excerpts of documents in support of projects reported
for the 2015 Assessment. These documents make clear that the water industry’s approach to understanding
assets and asset condition, as well as the industry’s approach to capital investment planning, are evolving and
yielding an increasingly robust understanding of long-term needs. In general, water utilities are developing
increasingly complete inventories of their assets. Some systems are adopting asset management techniques to
document the condition of assets and to plan for their rehabilitation or replacement. However, for many other
systems, the information and documentation provided indicates that a significant gap still exists between their
inventory of infrastructure and their knowledge of that infrastructure’s condition or remaining useful life.
13
Chapter 1: Findings - National Need
2015 Total National Need
The 20-year national infrastructure need estimated by the 2015 Assessment is $472.6 billion. The breakout of the
need by system size and type is presented in Exhibit 1.1.
Exhibit 1.1: Total National 20-year Need (in billions of January 2015 dollars)
System Size/Type
Need
Population
Served
(millions)***
Large Community Water Systems
(serving over 100,000 people)*
$174.4
141.7
Medium Community Water Systems
(serving 3,301 to 100,000 people)*
$210.6
139.4
Small Community Water Systems
(serving 3,300 and fewer people)
$74.4
23.4
Not-for-Profit Noncommunity Water Systems
$5.1
Total State Need
$464.6
American Indian Water Systems
§
$3.1
1.04
Alaska Native Village Water Systems
§
$0.7
0.08
Costs Associated with Proposed and Recently
Promulgated Regulations**
$4.2
Total National Need
$472.6
Note: Numbers may not total due to rounding.
* "Large" and "Medium" community water systems are defined the same as for the 2007 and 2011
Assessments but are different than in the 2003 and previous Assessments. See Appendix A for more
information.
† Based on 2007 Assessment findings adjusted to 2015 dollars, an updated inventory, and updated cost
models.
‡ Based on 1999 Assessment findings adjusted to 2015 dollars. Population cannot be determined.
§ Based on 2011 Assessment findings adjusted to 2015 dollars and updated cost models.
** Needs associated with the Proposed Radon Rule taken from EPA economic analyses. Population cannot
be determined.
*** Population for state, American Indian, and Alaska Native Village systems was derived on 12/8/2016
from EPA’s SDWIS Federal Reports website: https://ofmpub.epa.gov/apex/sfdw/f?p=108:1:::NO:1. The
information is for community water systems. These values do not include populations for systems defined
as “Federal Systems,” which are not eligible for DWSRF funding.
The Assessment addressed community water systems (CWSs) and noncommunity water systems (NCWSs) that
serve not-for-profit entities (referred to as not-for-profit noncommunity systems or NPNCWS) because only
14
those entities are eligible for DWSRF funding.
3
The results for large and medium CWSs, regulated by the states,
were derived from surveying all large systems in the United States (serving over 100,000 people) and a
statistically representative sample of the nation’s medium systems (serving between 3,300 and 100,000 people).
The results for small CWSs (serving fewer than 3,300 people), regulated by the states, Puerto Rico and U.S.
territories, were estimated from the Assessment conducted in 2007. The results for American Indian and Alaska
Native Village CWSs (primarily in the small system category, with the remainder in the medium system category)
were estimated from the Assessment conducted in 2011. The results for the NPNCWSs in states, Puerto Rico, and
U.S. territories were adjusted to 2015 dollars from the 1999 Assessment findings.
4
The need reported in the Assessment includes projects for
expanding, replacing, or rehabilitating existing infrastructure. It
also includes projects to construct new infrastructure in order to
preserve the physical integrity of water systems, increase
resiliency or efficiency, and to convey drinking water to existing
residential, commercial, and industrial customers. Projects vary
greatly in scale, complexity, and costfrom rehabilitating a
small storage tank to replacing an entire treatment plant to
constructing a high-capacity pipeline.
The results presented in this report will determine the allocation
of DWSRF capitalization grants and factor into the allocation of
the tribal set-aside funding to EPA regional offices for federal
fiscal years 2018 through 2021. Projects that are ineligible for
DWSRF funding, even though these necessary projects may
constitute significant capital expenditures for local
communities, are not included. The approach and
methodologies for discerning needs are further detailed in
Appendix A. A summary of the types of projects included in the Assessment, as well as specific types of
unallowable projects, is presented in Appendix C. EPA recognizes that projects not eligible for DWSRF funding
can be significant, if not critical, water system needs, but they are outside the scope of this Assessment. In
addition, the Assessment does not seek to capture information on the financing alternatives being pursued or
considered by systems for individual projects. The DWSRF is intended as a supplement to, not a replacement for,
funding by states, localities and rate payers.
3
The Assessment does not address federally-owned water systems as they are not eligible for DWSRF funding.
4
The confidence in the estimated need of NPNCWS is lower than that of the community water systems because it is based on data from
a small national sample of systems collected in 1999. However, the uncertainty this introduces is small because of the relatively small
need associated with NPNCWSs.
A community water system is a public water
system that serves at least 15 connections
used by year-round residents or that
regularly serves at least 25 residents year-
round. Cities, towns, and small communities
as well as mobile home parks and home-
owner associations may all own and operate
a community water system.
A noncommunity water system is a public
water system that is not a community water
system and that serves a nonresidential
population of at least 25 individuals daily for
at least 60 days of the year. Schools and
churches are examples of noncommunity
water systems.
15
2015 Total National Need Compared to EPAs Previous Assessments
EPA conducted its previous Assessments in 1995, 1999, 2003, 2007 and 2011. Exhibit 1.2 presents the total
national need for each Assessment. The first row shows the total need as reported in each Report to Congress,
i.e., “current year dollars.” These amounts are then adjusted for inflation (“cost adjustment factors” presented in
the exhibit) to report the need in 2015 dollars in the “Total National Need in 2015 Dollars” row in the exhibit.
This allows for a comparison in real dollars of the changes in needs as assessed by the survey efforts. EPA used
the Construction Cost Index (CCI) compiled by McGraw Hill Construction to adjust for inflation because it
includes adjustments for labor rates as well as the cost of materials.
Exhibit 1.2 also shows the percent change in the total national need between each assessment over the course of
the six national assessments. The 1.1 percent decrease in total need between 1995 and 1999 reflects refinements
in project acceptance criteria; particularly the exclusion of raw water dams and reservoirs from the Assessment
findings. The need increased substantially between the 1999 and 2003 Assessments due, in part, to the addition
in the 2003 Assessment of an asset inventory-based approach to better capture long-term replacement and
rehabilitation needs for existing infrastructure (as described in Appendix B). Between the 2003 and 2007
Assessments and the 2007 and 2011 Assessments, the need increased only 1.0 and 1.2 percent, respectively.
The 2015 findings, however, show a 10.3 percent ($44 billion) increase in need compared to the 2011 findings.
Exhibit 1.2: Total National 20-year Need Comparison to Previous DWINSA Findings (dollars in billions)
1995
1999
2003
2007
2011
2015
Total National Need
(as listed in Assessment Year's Report to
Congress in Current Year Dollars)
$138.4
$150.9
$276.8
$334.8
$384.2
$472.6
Cost adjustment factor to January 2015 dollars
(based on Construction Cost Index)
83%
66%
52%
27%
12%
--
Total National Need in 2015 Dollars*
$253.6
$250.9
$419.4
$423.7
$428.6
$472.6
Percent Change from Previous Assessment*
--
-1.1%
67.2%
1.0%
1.2%
10.3%
* Numbers may not total due to rounding.
Economic Benefits of Water Infrastructure Investment
The $472.6 billion represents the need associated with hundreds of thousands of miles of pipe, thousands of
treatment plant and source water projects, and billions of gallons of storage (based on survey responses).
Investments in water systems not only provide assurances of continued delivery of safe drinking water to
American homes, schools, and places of business, they are key to local economies across the United States.
The Department of Commerce Bureau of Economic Analysis (BEA) estimates that for each additional dollar
of revenue in the water and sewer industry, the increase in revenue that occurs in all industries is $2.62 in that
year. Further, adding one job in water and sewer creates 3.68 jobs in the national economy to support that job.
The estimate is based on a 2008 analysis done by the U.S. Conference of Mayors using BEA’s Regional Input-
Output Modeling System (RIMS II).
*
*
U.S. Conference of Mayors. “Local Government Investment in Municipal Water and Sewer Infrastructure:
Adding Value to the National Economy” (August 14, 2008).
16
Understanding the Increase in Need
The $44 billion increase in the total national need between the 2011 and 2015 Assessments is primarily due to
changes in needs reported for large and medium systems in fully participating states, which increased $42.2
billion. (See Chapter 2 and Appendix A for more information about full- versus partial-participation states.)
Combined, the changes in need for small systems, NPNCWSs, Alaska Native Village (ANV), and American
Indian (AI) systems and medium systems in partial-participating states represent only $1.7 billion of the increase.
As shown in Exhibit 1.3, between 2011 and 2015, the need reported for the rehabilitation, replacement and
upgrade of existing infrastructure increased by $53.7 billion (an increase of 18 percent) while new infrastructure
need for these systems actually declined by $11.4 billion, a decline of almost 30 percent. The dominant need to
replace or rehabilitate water systems’ existing infrastructure continues to increase relative to the needs for
investing in new infrastructure and is the cause of the overall increase in total national needs in 2015.
Most of the increase in existing infrastructure need for large and medium systems between 2011 and 2015 is for
replacement and rehabilitation of distribution and transmission pipe, which increased by $37.8 billion.
Distribution and transmission needs account for most of the nation’s needs, as described later in this section. On
a percentage basis, existing distribution and transmission need increased more than other category types. Based
on the information gathered during this Assessment, the increased needs are due to changes related to the survey
methodology for 2015 that allowed states to build on their efforts from the previous survey and improve the
estimate of their needs inventory.
Exhibit 1.3. Change in Need for New versus Existing Infrastructure between 2007, 2011 and 2015
Assessments*
Total Need
(billions of 2015 dollars)
Change in Need
(billions of 2015 dollars)
New vs. Existing
Infrastructure
2007
2011
2015
2007 to 2011
2011 to 2015
New Infrastructure
$63.1
$38.6
$27.2
-$24.4(-38.8%)
-$11.4(-29.5%)
Rehabilitation/Replace
ment/Upgrade of Existing
Infrastructure
$255.8
$291.4
$345.1
$35.7(+13.9%)
$53.7(+18.4%)
Net Increase
$318.8
$330.1
$372.3
$11.2(3.5%)
$42.2(+12.8%)
* Values are for large systems and medium systems in full-participation states. Excludes medium systems in partial-participation states, small systems,
NPNCWS, AI and ANV systems.
† Totals may not total due to rounding.
Exhibit 1.4 addresses the net increase in need from the perspective of system size. Between 2011 and 2015,
increase in replacement and rehabilitation needs was mostly attributed to medium systems ($33.0 billion out of
$53.7 billion). Whereas between 2007 and 2011, the increase was mostly attributed to large systems ($26.7 billion
out of $35.7 billion).
17
Exhibit 1.4. Change in Need by Medium versus Large Systems for Existing Infrastructure between 2007,
2011 and 2015 Assessments*
System Size
Change in Need (billions
of 2015 dollars)
2007-11
2011-15
Medium Systems
$8.9
$33.0
Large Systems
$26.7
$20.7
Net Increase
$35.7
$53.7
*
Values are for large systems and medium systems in full-participation states. Excludes medium systems in partial-participation states, small systems,
NPNCWS, AI and ANV systems.
Totals may not total due to rounding.
Changes in Survey Approaches
Since the 2003 Assessment, EPA has worked with the states to improve estimates of long-term needs for
infrastructure rehabilitation and replacement through enhanced accounting of water system assets. Some states
have reported that they made this a primary effort for the 2015 Assessment. In addition, for the 2015 Assessment,
EPA made use of a modified panel approach to resample a larger percentage of medium systems that participated
in the 2011 Assessment (see Appendix A for more details on this approach). For these systems, project data from
the 2011 Assessment were provided to water systems as a starting point for their 2015 survey response. By
allowing water systems to update their responses from the 2011 Assessment, these systems were able to build on
their efforts from the previous survey and improve the estimate of their existing assets. Some states also enhanced
their support for systems that participated in the 2015 Assessment. State-wide approaches, combined with the
improved systems’ responses, may have contributed to the increase in specific states’ need as well as the total
national need.
Column installation of a new concrete water storage tank in
Sleepy Hollow, NY.
18
State Approaches to Data Collection for the 2015 Assessment
“MassDEP increased its level of technical assistance to the surveyed systems for the 2015 survey. The
expanded technical assistance helped the local systems more completely describe their infrastructure
inventory and needs.”
-Steve McCurdy, MassDEP
“In Colorado, we revised our approach in 2015 by focusing closely on existing inventory. We started with
each system’s actual database of pipe inventory, tanks, pump stations etc…. The panel approach was helpful
because it allowed us to be more accurate and compare the different years of the survey. For instance, in
2015 we noticed two large systems with significant raw water pump station and pipeline networks that
delivered water from the mountains to the treatment facilities. These were not included in 2011 survey because
the systems did not realize that it should have been accounted for."
-Mark Henderson, CDPHE
"For the 2015 survey, North Carolina dedicated one full-time staff to the survey, with 25 regional engineers
also contributing to the effort. NC arranged training for these staff with special emphasis on the preferred
‘inventory-based’ approach, which ensured that systems fully captured the needs associated with all existing
infrastructure. In 2011, NC found that one water system with 9 million feet of pipe listed only 900,000 feet in
the inventory. In another city, the 2011 DWINSA included only one pipe replacement project of 17,000 feet.
In 2015, NC used an inventory-based approach to document that city’s more complete pipe replacement needs
of 37,000 feet. With a better understanding of the inventory-based approach and better QA/QC, the 2015
survey results better reflect the state’s needs."
-Amanjit Paintal, NC DEQ
“In 2011, TDEC sent the surveys to the water system with instruction on filling out the forms. We found that
there was some confusion and inconsistency in how the systems reported information. Due to the inconsistency
in the previous report, in 2015, TDEC staff contacted the systems and worked with them over the phone to
complete the surveys. We believe this lead to more accurate reporting.”
-Anna Sartors, Tennessee DEC
19
Exhibit 1.5 provides a summary of needs across all Assessments since 1995 by system size and type. It also
includes needs identified in this and past Assessments for regulations that are or were proposed or recently
promulgated at the time of the assessment.
Exhibit 1.5: Total 20-Year Need by System Size/Type (and Regulation) for Each Assessment (in billions
of January 2015 dollars)
System Type
(and Regulation)
1995
1999
2003
2007
2011
2015
Large CWSs (serving more than
100,000 people)
$147.2
$161.9
$174.4
Medium CWSs (serving 3,301 to
100,000 people)
$183.6
$180.5
$210.6
Combined Medium/Large CWSs*
$183.0
$174.7
$342.3
Small CWSs (serving 3,300 or
fewer people)
$68.2
$51.8
$51.8
$75.2
$72.0
$74.4
Not-for-Profit Noncommunity
Systems
$5.1
$5.1
$5.1
$5.1
$5.1
American Indian Water Systems
$2.4
$1.9
$1.9
$1.9
$3.0
$3.1
Alaska Native Village Water
Systems
$1.8
$1.8
$1.8
$0.7
$0.7
Proposed and Recently
Promulgated Regulations
$15.5
$16.5
$8.8
$5.5
$4.2
Total National Need
$253.6
$250.9
$419.4
$423.7
$428.6
$472.6
Note: Numbers may not total due to rounding.
*For 1995, 1999, and 2003 Assessments, medium and large CWS need is combined due to changes in the definitions of the
system sizes.
Water distribution system improvement project in Williamsburg, CO.
20
EPA’s Assessment continues to estimate a need within the range identified in other important reports. Below is a
summary of the findings from other studies, all adjusted by EPA to January 2015 dollars.
The 2002 Congressional Budget Office report Future Investment in Drinking Water and Wastewater
Infrastructure,” estimates annual water system needs of $18.5 billion to $31.9 billion. This extrapolates to
a 20-year need in the range of $369.5 to $637.8 billion.
5
The Water Infrastructure Network’s “Clean and Safe Water for the 21st Century - A Renewed National
Commitment to Water and Wastewater Infrastructure,” estimates water system needs of $31.8 billion
annually. This extrapolates to $636.4 billion over 20 years.
6
The 2012 American Water Works Association report “Buried No Longer: Confronting America’s Water
Infrastructure Challenge” estimated at least $1 trillion will be required over a 25-year period from 2010
through 2035 in order to restore existing water system pipe that has reached the end of its useful life and
to expand pipe networks to meet growing populations. This estimate is significantly higher than the
transmission and distribution total for EPA’s 2015 Assessment, in part because it includes drinking water
infrastructure investment needs related to population growth (which are not eligible needs under the
DWSRF), and covers a longer period of time.
7
Total National Need by Project Category
Infrastructure needs of water systems can be grouped into four major categories based on project type. These
categories are drinking water source, transmission and distribution, treatment, and storage. Infrastructure in each
category fulfills an important function in delivering safe drinking water to the public. Most needs were assigned
to one of these categories. An additional “other” category is composed of projects that do not fit into one of the
four, such as system-wide supervisory control and data acquisition (SCADA) or emergency generators. The
treatment category includes most of the need that is associated with drinking water regulations. Exhibit 1.6 shows
the total national need by project category. Exhibit 1.7 shows the total national need by water system size and
type, as well as by project category.
5
Congressional Budget Office, Future Investment in Drinking Water and Wastewater Infrastructure” (November, 2002), p. ix. Needs
were reported in 2001 dollars and have been adjusted to January 2015 dollars for comparison purposes.
6
Water Infrastructure Network, “Clean and Safe Water for the 21st Century - A Renewed National Commitment to Water and
Wastewater Infrastructure” (undated), p. 3-1. Needs were assumed to be in 1999 dollars based on the planning period and data used.
Needs have been adjusted to January 2015 dollars for comparison purposes.
7
American Water Works Association “Buried No Longer: Confronting America’s Water Infrastructure Challenge” (February 2012),
p. 9. http://www.awwa.org/Portals/0/files/legreg/documents/BuriedNoLonger.pdf (Needs were reported in 2010 dollars and have been
adjusted to January 2015 dollars for comparison).
21
Exhibit 1.6: Total 20-year Need by Project Category (in billions of January 2015 dollars)
Note: Numbers may not total due to rounding.
Exhibit 1.7: Total 20-Year Need by System Size/Type and Project Category (in billions of January 2015
dollars)
System Size/Type
Distribution
and
Transmission
Treatment
Storage
Source
Other
Total Need
Large Community Water
Systems (serving over
100,000 people)*
$120.1
$32.5
$13.4
$6.1
$2.3
$174.4
Medium Community Water
Systems (serving 3,301 to
100,000 people)*
$144.9
$33.7
$20.0
$7.8
$4.3
$210.6
Small Community Water
Systems (serving 3,300 and
fewer people)
$44.6
$11.0
$11.3
$6.7
$0.9
$74.4
Not-for-Profit
Noncommunity Water
Systems
$0.6
$1.0
$2.5
$1.0
$0.0
$5.1
Total States and U.S.
Territories Need
$310.2
$78.2
$47.1
$21.6
$7.4
$464.6
American Indian and Alaska
Native Village Systems
§
$2.4
$0.6
$0.5
$0.2
$0.1
$3.8
Costs Associated with
Proposed and Recently
Promulgated Regulations**
$4.2
$4.2
Total National Need
$312.6
$83.0
$47.6
$21.8
$7.5
$472.6
Note: Numbers may not total due to rounding.
*
“Large” and “Medium” community water system population ranges were the same for the 2007, 2011, and 2015 Assessments but
differed in the 2003 and previous Assessments. See Appendix A for more information.
† Based on 2007 Assessment findings adjusted to 2015 dollars, an updated inventory, and updated cost models.
‡ Based on 1999 Assessment findings adjusted to 2015 dollars.
§ Based on 2011 Assessment findings adjusted to 2015 dollars and updated cost models.
** Needs associated with the Proposed Radon Rule - taken from EPA economic analyses.
22
Exhibit 1.8 shows the changes in need by project category over the last two decades across all six Assessments
from 1995 to 2015.
Exhibit 1.8: Total 20-Year Need by Project Category for Each Assessment (in billions of January 2015
dollars)
Transmission and Distribution Needs
Transmission and distribution projects are the largest project category, totaling $312.6 billion over the next 20
years (66 percent of the total need in 2015). This category includes projects for rehabilitation and replacement of
existing water mains, installing new pipe to eliminate dead end mains and the resulting stagnant water, installing
new mains in areas where existing homes do not have a safe and adequate water supply, and installing or
rehabilitating pumping stations to maintain adequate pressure. It also includes projects to address the replacement
of appurtenances, such as meters to record flow and water consumption, backflow-prevention devices to avoid
contamination, and valves that are essential for controlling flows and isolating problem areas during repairs.
Although the least visible component of a public water system, the buried pipes of a transmission and distribution
network generally account for most of a system’s capital value. Even small rural systems may have several
hundred miles of pipe to reach all customers. In larger cities, replacement or rehabilitation of even small segments
of the extensive underground networks of water pipes can be costly, approaching hundreds or thousands of dollars
per foot (based on survey responses). The costs to a city include cost of construction as well as the costs related
to disruption to the city’s commerce. Regardless of water system size, projects addressing water mains and related
1995 1999 2003 2007 2011
Assessment
Distribution/
Transmission
$50
$0
Treatment
$150
$100
Storage
$200
Source
$250
Other
$500
$450
$400
$350
$300
Billions
of Dollars
23
infrastructure present challenges. Pipe projects are typically driven by a utility’s need to maintain drinking water
quality and pressure as it travels through the distribution system from the treatment plant to the tap.
A substantial portion of the transmission and distribution need is for replacing or refurbishing aging or
deteriorating transmission and distribution mains. These projects are critical to the delivery of safe drinking water
and can help ensure compliance with many regulatory requirements. Failures in transmission and distribution
mains can interrupt the delivery of water and introduce dangerous contaminants into the drinking water supply.
The rate at which water mains require replacement or rehabilitation varies greatly by pipe material, age, soil
characteristics, weather conditions, and construction methods. Systems that have not been able to implement or
prioritize adequate programs to rehabilitate or replace mains may have proportionally more aged infrastructure,
and therefore a higher level of need.
Total Length of Water Mains in the United States
A water system’s inventory of pipe is typically one of its largest and most important assets. It is mostly buried,
hidden from direct condition assessment, and expensive to replace. Estimates of how long pipes will last are
critical to assessments of a system’s infrastructure investment needs.
EPA estimates that community water systems currently have a total of 2.2 million miles of transmission lines and
distribution mains. This estimate is based on data provided for medium and large systems in the 2015 Assessment
and data provided for small systems in the 2007 combined DWINSA and Community Water System Survey. Note
that this is the total amount of pipe, not the amount in need of replacement. EPA is able to report this information
for the first time because a high percentage of respondents for the 2015 Assessment (98 percent) provided EPA
with a total amount of pipe in their system.
Exhibit 1.9 shows pipe length in miles by system type and population category as well as the average length of
pipe per system by size category. Large systems, defined as those serving over 100,000 people, have
approximately 607,400 miles of pipe and average 943 miles per system (based on a census of large systems).
Medium-sized systems, those serving between 3,301 and 100,000 people, have approximately 1,234,300 miles of
pipe and 134 miles per system (based on a statistical sample of systems at the state level). Small systems, those
serving fewer than 3,301 people, have approximately 379,600 miles of pipe and average 9.5 miles per system
(based on a national sample).
American Indian systems have an approximate total of 9,930 miles of pipe and average 13.5 miles per system.
This relatively high amount compared to small systems regulated by the states may be due to the rural nature of
many Native American communities. Alaska Native Village systems have an approximate total of 480 miles with
just 2.9 miles per system. This low value is because many Alaska Native Villages do not have piped water due to
permafrost conditions. American Indian and Alaska Native estimates are based on a statistical sample of systems.
24
Exhibit 1.9. Total Length of Pipe for CWSs in the Nation by System Size (Year 2015)
System Type
Population Category
Total Length of Distribution
and Transmission Mains
(in miles)*
Average Length of
Distribution and Transmission
Mains per System
(in miles)
State Systems
Large CWSs
607,400
943.2
50,000-100,000
235,800
346.4
10,001-50,000
617,100
162.9
3,301-10,000
381,400
80.6
Subtotal, Medium CWSs
1,234,300
134.1
501-3,300
271,600
21.1
101-500
65,500
4.4
25-100
42,500
3.5
Subtotal, Small CWSs
379,600
9.5
American
Indian (AI)
Systems
10,001-50,000
820
51.0
3,301-10,000
4,220
74.0
501-3,300
3,980
15.1
101-500
800
2.7
25-100
110
1.1
Subtotal, AI CWSs
9,930
13.5
Alaska Native
Village (ANV)
Systems
3,301-10,000
100
26.0
501-3,300
160
4.6
101-500
190
2.0
25-100
30
1.0
Subtotal, ANV CWSs
480
2.9
National Total
2,231,710
44.1
Note: Exhibit includes medium system pipe length in partial participation states. See Appendix A for a discussion regarding partial
participation states.
*
Numbers may not total due to rounding.
Approaches to Reporting Pipe Rehabilitation/Replacement Need
Beginning with the 2007 Assessment, EPA and the DWINSA Workgroup of state and EPA regional office
representatives set a policy specific to reporting pipe rehabilitation or replacement 20-year needs. The policy
allows a system to report the rehabilitation and/or replacement of up to 10 percent of its existing water mains over
the 20-year period (0.5 percent annually) based on a simple statement of need. This default amount was not based
on an EPA recommended replacement rate, but rather was based on the average pipe replacement rate in the U.S.
If a system had 20-year water main rehabilitation/replacement needs greater than 10 percent of their existing
mains, the projects would be included if independent documentation of this need was provided. The
documentation could include budget and planning documents such as capital improvement plans, master plans,
and project-specific design documents. Asset management plan documents could also be used if the plan has been
through the critical step of water system review of asset management program outputs. Note that this pipe
replacement rate assumes that pipe that is placed in the ground today will have a useful life of 200 years.
25
Exhibit 1.10 below presents the percentage of survey respondents that made use of the default approach and the
percent of respondents that took other approaches (reported less than the default 10 percent or submitted
independent documentation to support their pipe rehab/replacement needs). Exhibit 1.11 presents the percentage
of need correlated to each approach.
Exhibit 1.10: System Approaches for Pipe
Rehab/Replace/Replacement by Percentage of
Exhibit 1.11: Pipe Rehab/Replace Need by
System Approach
Note: Numbers may not total due to rounding.
60-inch transmission main for a project in Oak Lawn, IL.
26
Lead Service Lines
Lead service lines have become an increasing priority as issues in Flint, Michigan, have gained the nation’s
attention. One way lead can enter drinking water is when plumbing materials, including service pipes that
contain lead corrode, especially where the water has high acidity or low mineral content.
In 1991, EPA published the Final Regulatory Impact Analysis of National Primary Drinking Water
Regulations for Lead and Copper, which estimated that there were 10.2 million lead service lines in the
country at that time (USEPA 1991)
*
. More recently, an April 2016 report in the Journal of the American
Water Works Association estimated that there are currently 6.1 million lead service lines in the U.S.
**
The DWINSA requests that systems identify capital infrastructure needs, which could include needs related
to replacing lead service lines. However, for the purposes of the Assessment, water systems are not
specifically asked to report the total number of lead service lines in their system. Based on data from large
and medium systems in the 2015 Assessment and from small systems in the 2007 Assessment, water
systems identified needs for replacement of approximately 1.4 million lead service lines over the 20-year
period of January 2015 through December 2034. The estimated total cost of replacing these lead service
lines is $4.2 billion in 2015 dollars.
Several factors likely contribute to the DWINSA estimate of the number of lead service lines needing
replacement, none of which reflect EPA’s perspective on the importance of this activity:
Systems that may have lead lines but have not been experiencing lead and copper action level exceedances
may not report a need for replacements.
Systems that control lead in their drinking water through corrosion control measures may not report a need
for replacements.
Many systems address lead service lines as they are encountered in the process of rehabilitating or
replacing water mains. Thus, systems may not know in advance how many lead service lines they will be
replacing.
For purposes of the needs assessment, water systems may not focus on lead service line replacement.
In the absence of project-specific data, the Assessment models the cost of lead service line replacement at
$3,777/line. The cost model was derived from actual lead service line cost data submitted by water systems
in past survey responses. While small compared to other needs, lead service line replacement is still a
priority for many systems, and may be an increasing priority in future needs assessments as a result of
heightened awareness about lead and drinking water issues.
EPA continues to work on finalizing the revisions to the Lead and Copper Rule.
*EPA. “Final Regulatory Impact Analysis of National Primary Drinking Water Regulations for Lead and
Copper” (1991).
**AWWA. Journal of the American Water Works Association. “National Survey of Lead Service Line
Occurrence” (Cornwell, D; Brown, A; and Via, S). (April 2016).
27
Treatment Needs
The total 20-year national need for treatment is estimated to be $83.0 billion. This category includes the
construction, expansion, and rehabilitation of facilities to reduce contamination through treatment processes. As
discussed in the Section “Needs Associated with SDWA Regulations,” a large percentage of the regulatory need
is in this category. Drinking water treatment facilities vary significantly depending on the quality of their source
water and type of contamination addressed. Treatment systems range from a simple chlorinator for disinfection
to a complete conventional treatment system with sedimentation, filtration, disinfection, laboratory facilities,
waste handling, and computer automated monitoring and control devices.
The treatment category also includes projects to remove contaminants that adversely affect the taste, odor, and
color of drinking water. Treatment for these “secondary contaminants” often involves softening the water to
reduce magnesium and calcium levels, or applying sequestrants to chemically bind iron and manganese in order
to prevent fixture discoloration, taste, or other aesthetic issues. Although not a public health concern, the aesthetic
problems caused by secondary contaminants may prompt some consumers to seek more palatable, but less safe
or affordable sources of water.
Increasingly, many surface water treatment systems are installing advanced treatment processes such as
membrane filters and advanced oxidation to achieve drinking water quality objectives. The movement to more
advanced treatment is driven by a number of factors, including customer expectations for drinking water quality;
state and local requirements; or a system’s own operational benchmarks.
Source Needs
The total 20-year national need for source water infrastructure is estimated at $21.8 billion. Source water needs
include construction or rehabilitation of surface water intake structures, drilled wells, and spring collectors. Needs
for dams and raw water reservoirs are not eligible for DWSRF funding and are therefore excluded from all
Assessments since 1995 (the 1995 Assessment was completed prior to establishment of the DWSRF and its
policies).
Drinking water may come from either ground water or surface water sources, and the treatment needs for removal
of contaminants from these sources can differ considerably. A high-quality source water can minimize the
possibility of microbial or chemical contamination and may require less treatment. Many source water needs
involve construction of new surface water intake structures or drilling new wells to obtain higher-quality raw
water.
A water source should provide an adequate supply to enable the water system to maintain minimum pressures.
Low water pressure may result in the intrusion of contaminants into the distribution system. The 2015 Assessment
includes a number of projects to expand the capacity of intake structures and add new wells to address supply
deficiencies facing existing customers.
Storage Needs
The 20-year national need estimated for storage projects is $47.6 billion. This category includes projects to
construct, rehabilitate, or cover finished water storage tanks, but it excludes dams and raw water reservoirs (unless
the raw water basins are located at the treatment facility and are part of the treatment process) because the DWSRF
regulations specifically exclude these projects from DWSRF funding. It is critical that water systems have
sufficient storage to provide adequate supplies of treated water to the public, particularly during periods of peak
28
demand. Sufficient storage enables the system to maintain the minimum pressure required throughout the
distribution system to prevent the intrusion of contaminants into the distribution network.
Other Needs
DWSRF-eligible projects that are not included in the previous four categories are grouped as “other” needs. These
needs account for $7.5 billion of the total 20-year national need. Examples of “other” projects are SCADA systems
for monitoring and controlling water system facility operations and emergency generators that were not assigned
to another category.
New Infrastructure Needs versus Needs Associated with Existing Infrastructure
As demonstrated in Exhibit 1.12, at a national level, the needs associated with the rehabilitation, upgrade,
expansion, or replacement of existing infrastructure are much larger than the DWSRF-eligible needs associated
with the construction of new infrastructure (93 percent compared to 7 percent).
Exhibit 1.12: Percent of Large and Medium System Need by New vs. Existing Infrastructure and by
Project Category*
Existing Infrastructure
(Rehabilitation,
Expansion/Upgrade,
Replacement)
New
Infrastructure
(DWSRF-eligible Only)
Source
75%
25%
Treatment
91%
9%
Storage
90%
10%
Pipe
94%
6%
T&D Needs other than Pipe
96%
4%
Other Needs
87%
13%
Overall
93%
7%
* Percentages represent the needs of large systems and medium systems surveyed in 2015. It does not include the needs of medium
systems in partial participation states (discussed in Appendix A), small systems, not-for-profit noncommunity systems, American
Indian systems, or Alaska Native Village systems.
Transmission and Distribution (T&D) needs other than pipe include infrastructure such as pump stations, valves, meters, backflow
prevention, and service lines.
Includes infrastructure such as system-wide SCADA or emergency generators not specifically associated with another need category.
There is more existing infrastructure that must be rehabilitated or replaced to maintain their reliable function
over the 20-year Assessment period than new infrastructure needed to address a deficiency. In addition, the
Assessment does not include needs for new infrastructure projects that are driven by needs that are not
DWSRF-eligible, such as anticipated population growth.
Projects for rehabilitation, replacement, upgrade, or expansion are commonly intended to address age and
deterioration of a water system’s existing infrastructure. Projects for new infrastructure are generally driven by a
29
need for system expansion such as a new storage tank to address pressure deficiencies or a new water main to
loop water distribution system dead end lines to reduce stagnation. Projects for new infrastructure to accommodate
anticipated growth or for fire suppression are not eligible for DWSRF funding and are not included in the
DWINSA estimate of need. The need for replacement or rehabilitation of water systems’ existing infrastructure
has continued to increase relative to the needs for new infrastructure, resulting in the overall increase in total
national needs in 2015.
It should be noted that the needs for individual systems and states can vary from the overall national findings.
For example, the survey results find that nationally, the need for new pipe is just 6 percent of the total pipe-
related need; however, one wholesale system documented a need for new pipe that represents 52 percent of its
total need. That system uses surface water as its primary source and reported a significant need to extend new
water transmission lines to neighboring communities where ground water withdrawals have caused subsidence
problems. This system’s need for new pipelines contributed substantially to the state’s high percentage of new
pipe need (15 percent of the total pipe need in the state). See Chapter 2 for more discussion regarding need at
the state level.
Need by System Size
Exhibit 1.13 shows the distribution of infrastructure need by community water system size and total population
served. As this exhibit shows, the very small number of large community water systems serve a large portion of
the U.S. population (46.4 percent) but account for a smaller percentage (37.6 percent) of the overall drinking
water infrastructure investment need, reflecting the economies-of-scale in large drinking water systems.
Conversely, the very large number of small community water systems serve only 7.8 percent of the U.S.
population, but account for a disproportionate percentage (16.5 percent) of the CWS need. Medium community
water systems represent the largest portion of the need (45.8 percent), and their need is proportional to the
population served (45.8 percent).
American Indian and Alaska Native Village communities are included in Exhibit 1.13. These systems serve
primarily small communities. For example, the 2011 Assessment found that approximately 90 percent of the 791
American Indian water systems served fewer than 3,300 people. Similarly, for Alaska Native Villages, in 2011
all but four of the 165 systems served 3,300 or fewer people and none served over 10,000 people.
30
Exhibit 1.13: Community Water System 20-year Need by Size and Population (in billions of January
2015 dollars)
System Size*
Need
Water Systems
Total Population Served
$ Billions
% of
Need
Number of
Systems
% of
Water
Systems
Population
(millions)
% of
Population
Served
§
Large Community Water
Systems (serving over 100,000
people)
$174.4
37.6%
644
1.3%
141.7
46.4%
Medium Community Water
Systems (serving 3,301 to
100,000 people)
$212.3
45.8%
9,279
18.7%
140.1
45.8%
Small Community Water
Systems (serving 3,300 and
fewer people)
$76.6
16.5%
39,482
79.9%
23.9
7.8%
Note: Percentages may not add to 100 due to rounding.
* This exhibit reports the needs for community water systems in the states, Washington D.C., Puerto Rico, the U.S. territories,
American Indian systems, and Alaska Native Village systems. It does not include findings for not-for-profit noncommunity water
systems.
Based on the 2015 DWINSA sample frame for state-regulated systems and the 2011 DWINSA sample frame for American Indian
and Alaska Native Village systems.
Population was derived on 12/8/2016 from EPA’s SDWIS Federal Reports website:
https://ofmpub.epa.gov/apex/sfdw/f?p=108:1:::NO:1. The information is for community water systems only and includes populations
for American Indian and Alaska Native Village water systems. It does not include populations for systems defined as “Federal
Systems,” which are not eligible for the DWSRF.
§
Values are a percentage of the population that is served by CWSs. Does not include homes served by other sources such as
individual wells or small multi-family supplies.
Needs Associated with SDWA Regulations
As shown in Exhibit 1.14, just over 12 percent of the total national drinking water infrastructure need ($57.6
billion), is related to compliance with the SDWA regulations. Projects that are directly attributable to specific
SDWA regulations are collectively referred to as the “regulatory need.” Many infrastructure improvements are to
ensure a reliable supply of safe water is provided and are not directly linked to a particular regulatory requirement,
although they facilitate to compliance. Most of the regulatory need involves the upgrade, replacement, or
installation of treatment technologies. Of the $57.6 billion of regulatory need, $41.8 billion is related to treatment;
while $11.0 billion, $2.4 billion, $2.4 billion, and $0.1 billion are related to the distribution and transmission,
storage, source, and ‘other’ categories, respectively.
31
Exhibit 1.14: Total Regulatory vs. Non-Regulatory 20-year Need (in billions of January 2015 dollars)
Note: Numbers may not total due to rounding.
The Assessment divides the regulatory need into microbial and chemical regulations. Exhibit 1.15 presents the
regulatory needs by these categories and compares the regulatory need with non-regulatory 20-year need. Note
that the need associated with the 1999 Proposed Radon Rule is included in the chemical category.
32
Exhibit 1.15: Total Regulatory Need by Microbial and Chemical Regulations vs. Non-Regulatory 20-year
Need (in billions of January 2015 dollars)
Note: Needs associated with the Proposed Radon Rule are included in chemical regulations.
Existing Regulatory Needs
The infrastructure needs associated with water systems complying with existing EPA regulations are identified
through the completed survey questionnaires. The reported infrastructure needs associated with existing
regulations are reflected in the needs previously presented by system size, project category, and total need.
Existing Regulations for Microbial Contaminants
The surface water treatment regulations (Surface Water Treatment Rule, Interim Enhanced Surface Water
Treatment Rule, Filter Backwash Recycling Rule, Long Term 1 Enhanced Surface Water Treatment Rule, and
Long Term 2 Enhanced Surface Water Treatment Rule), the Revised Total Coliform Rule, and the Ground Water
Rule are existing SDWA regulations that address microbial contamination. The Stage 1 and Stage 2
Disinfectants/Disinfection Byproducts Rules regulate the disinfectant and disinfection byproduct levels in
distribution systems and are grouped with the microbial rules.
Projects for compliance with microbial regulations account for 69 percent of the total regulatory need. Under
these regulations, systems using surface water sources must provide treatment to minimize microbial
contamination. In most cases, this means installing, upgrading, or rehabilitating treatment plants to control
dangerous human pathogens such as the bacterium E. coli, the virus Hepatitis A, and the protozoans Giardia
lamblia and Cryptosporidium.
Existing Regulations for Chemical Contaminants
This category includes regulations governing more than 80 inorganic or organic drinking water contaminants for
which infrastructure projects may be needed. This estimate includes projects intended to help systems comply
with the Nitrate/Nitrite Standard, the revised Arsenic Standard, the Lead and Copper Rule, and other regulations
33
that set maximum contaminant levels or treatment techniques for organic and inorganic chemicals. Examples of
these projects include installation of aerators to remove volatile organic compounds or ion exchange units to
remove inorganic contaminants from the water. This category also includes projects that are required to address
secondary standards such as iron, manganese, hardness, taste and odor. Examples of these projects include
oxidation/filtration to remove iron and/or manganese, softening to remove hardness, and distribution projects to
address water quality issues such as stagnant water.
Proposed or Recently Promulgated Regulatory Needs
In general, water systems can readily identify the infrastructure needs required for compliance with existing
regulations, but systems typically have not accounted for the infrastructure needed to comply with regulations
that are in proposed status or that have been recently promulgated at the time of data collection. Consequently,
for these needs, EPA derives the capital infrastructure estimates from the Economic Analysis (EA) that the
Agency published when proposing each regulation, or from the final EA if the regulation has been recently
promulgated. The only regulation that was in this category for the 2015 Assessment is the 1999 Proposed Radon
Rule. The total cost of complying with the Proposed Radon Rule, estimated to be $4.2 billion (in January 2015
dollars), is included in the Assessment as a regulatory need in the treatment and chemical contaminant categories.
Because the EAs rely on regional data, they are not appropriate predictors of state-specific needs. Therefore, the
costs associated with the Proposed Radon Rule are allocated at a national level, not apportioned to each state.
Well rehabilitation site in Peoria, AZ.
34
Asset Management
Understanding the Trade-Off between Capital and Operating Expenditures
The cost to the customer of safe drinking water provided by community water systems in the U.S. is typically
far lower than the water systems’ cost to provide that service to their customers. The true cost of providing
safe drinking water may be misunderstood, as much of the infrastructure is hidden from view. The typical
drinking water utility has seven times more asset value than annual operating income.
*
Customers expect a water utility to deliver high quality water, in as much quantity as needed, at adequate
pressures, and without interruption. Maintaining that level of service requires ongoing reinvestment to replace
infrastructure that has reached the end of its useful life. Avoiding such reinvestments will result in increasing
operation and maintenance expense or a significant increase in system failures.
Well managed utilities implement asset management programs to help them make prudent, economically
justified decisions regarding capital investment. The age of pipe alone is a poor indicator of the need for that
pipe to be replaced. Through asset management, utilities document the condition and failure history of their
piping network and other assets. Over time, these utilities can confidently predict the likely remaining useful
life in their assets. In some cases, utilities will find that they need to replace pipe sooner than age would
suggest, while in other cases, they will find that pipe can be expected to provide many more years of service
despite its age.
As a result of implementing asset management, some utilities are documenting larger capital investment needs
than they had previously anticipated. Others are finding that there is greater remaining useful life in their assets
than they had previously assumed. Deciding when to replace a given length of pipe ultimately depends upon
a utility’s target level of service and the risk the utility accepts of that particular section of pipe failing. The
target level of service for the entire utility may incorporate differing levels of failure risk for different
components of their distribution system. Pipe serving a hospital or other critical infrastructure may be managed
to a lower risk of failure than will pipe serving a commercial area.
Many utilities are only in the very early stages of developing an asset management program, as evidenced by
the reliance of most survey respondents on the survey’s baseline pipe replacement rate. That baseline rate of
0.5 percent per year, or 10 percent over 20 years, reflects the current documented rate of replacement of pipe
within the drinking water industry. A 0.5 percent per year replacement rate imputes a 200-year life to pipe.
*
Water Research Foundation and U.S. Environmental Protection Agency; Improving Water Utility Capital
Efficiency; 2009 Page 10
35
Chapter 2: Findings - State Need
State-Specific Needs
Since federal fiscal year 1998, the SDWA has
required EPA to allot DWSRF grants to each state
based on the findings of the most recent Assessment.
Given the critical role of this Assessment in
determining DWSRF capitalization grant allocations,
obtaining highly credible and statistically valid
estimates of each state’s need is essential. Exhibits 2.1
and 2.2 show the total DWSRF-eligible need for
states, Puerto Rico, the District of Columbia, and the
U.S. territories by project category and system size.
Exhibit 2.3 is a map indicating each state’s 20-year
total need.
DWSRF capitalization grants for fiscal years 2018
through 2021 will be allocated to states based on the
findings of the 2015 Assessment. The funding is
allocated by first setting aside a percentage allotment
to American Indian and Alaska Native Village water
systems (most recently set at 2.0 percent), and a
percent allotment to the U.S. territories, including the
U.S. Virgin Islands, Guam, the Commonwealth of the
Northern Mariana Islands, and American Samoa
(most recently 1.5 percent). The Assessment findings
are used to help divide these set-asides among these
entities. The remaining funds are then divided among
the states, Puerto Rico, and the District of Columbia,
based on the determination of each state’s relative
percentage of the total “state need,” with each
receiving no less than the 1 percent minimum
allotment.
States that received the minimum allotment of 1
percent in the most recent allocation were given the
option of a lower level of participation in the Assessment. As discussed below, only the large systems in these
states were surveyed, and data were not collected for medium sized systems. This option was provided to reduce
the burden on these states and allow for resources to be focused on the large systems. These states’ needs are
reported as one group referred to as “partial participation” states.
Partnership for Determining State Need
The substantial effort involved in collecting data and
calculating water systems’ 20-year needs relies on a
partnership between EPA, the states, and the utilities
themselves. Each partner makes a valuable
contribution to estimating the DWSRF-eligible needs
of drinking water systems.
Water System. Operators and managers of water
utilities have extensive on-the-ground knowledge of
their system’s infrastructure and condition. These
personnel are in the best position to assess their
infrastructure needs.
States. State personnel have considerable knowledge
of the systems in their state, and states have the staff
that are trained to assist systems in completing this
Assessment. As DWINSA workgroup members, the
states also work with EPA towards consensus
development of Assessment policies and methods to
ensure consistency across the states.
EPA. EPA’s primary roles are to serve as the quality
assurance agent for the data collection effort, to ensure
that survey policies and methodologies are met, and to
serve as a technical resource to assist with capturing
complete and accurate 20-year needs. EPA provides
oversight for survey submittals to encourage full
reporting, to ensure consistency and fairness between
states, and to control for any state bias.
Exhibit 2.1: State 20-year Need Reported by Project Category (in millions of January 2015 dollars)
* The needs for partial participation states (discussed in Appendix A) are presented cumulatively and not by state. The list of 14
partial participation states is shown in Exhibit 2.4.
36
State
Transmission and
Distribution
Treatment
Storage
Source
Other
Total
Alabama
$8,942.6
$1,097.5
$815.8
$193.6
$212.2
$11,261.9
Arizona
$5,837.1
$1,655.2
$1,042.2
$393.6
$202.4
$9,130.5
Arkansas
$5,461.2
$900.0
$643.4
$216.5
$155.7
$7,376.8
California
$31,685.9
$9,199.4
$6,967.2
$2,565.4
$615.2
$51,033.2
Colorado
$6,166.4
$2,722.4
$924.3
$228.4
$147.1
$10,188.7
Connecticut
$2,542.1
$770.4
$400.9
$187.6
$116.7
$4,017.7
District of Columbia
$1,573.8
$62.9
$104.5
$0.0
$0.7
$1,741.9
Florida
$13,734.0
$4,702.5
$1,551.6
$1,446.2
$452.2
$21,886.4
Georgia
$9,362.6
$1,676.8
$973.7
$365.4
$81.7
$12,460.1
Illinois
$13,494.8
$3,537.2
$1,751.0
$1,709.1
$418.2
$20,910.4
Indiana
$5,056.2
$1,198.1
$704.6
$430.2
$131.1
$7,520.2
Iowa
$5,858.5
$945.5
$607.1
$376.2
$65.1
$7,852.4
Kansas
$3,727.9
$870.5
$442.0
$235.2
$47.1
$5,322.6
Kentucky
$6,320.7
$929.7
$648.8
$206.7
$126.2
$8,232.0
Louisiana
$5,117.5
$1,064.7
$644.9
$333.0
$170.5
$7,330.6
Maine
$883.3
$198.2
$181.6
$72.6
$12.7
$1,348.3
Maryland
$6,959.4
$1,243.2
$810.1
$271.2
$46.1
$9,330.1
Massachusetts
$8,601.4
$1,850.6
$1,184.3
$330.7
$277.4
$12,244.4
Michigan
$9,084.7
$2,130.7
$1,042.0
$553.8
$235.0
$13,046.2
Minnesota
$4,416.6
$1,398.7
$912.3
$581.9
$198.4
$7,507.9
Mississippi
$3,090.2
$775.5
$561.2
$331.9
$64.4
$4,823.2
Missouri
$6,302.2
$1,305.1
$907.0
$374.3
$30.2
$8,918.9
Nevada
$2,844.4
$1,083.6
$454.3
$894.3
$40.2
$5,316.8
New Jersey
$5,381.1
$1,774.3
$888.5
$413.1
$126.1
$8,583.1
New York
$14,643.4
$3,974.7
$2,835.0
$1,035.2
$277.6
$22,765.9
North Carolina
$11,803.7
$2,453.2
$1,327.7
$786.7
$350.9
$16,722.2
Ohio
$8,970.4
$2,330.8
$1,258.5
$571.8
$273.9
$13,405.4
Oklahoma
$4,805.1
$1,043.4
$625.4
$258.7
$126.3
$6,858.9
Oregon
$3,742.4
$1,084.7
$1,022.2
$298.1
$103.0
$6,250.4
Pennsylvania
$11,134.2
$2,787.9
$1,881.9
$617.2
$350.3
$16,771.6
Puerto Rico
$2,201.9
$921.1
$397.6
$124.1
$59.5
$3,704.2
South Carolina
$4,555.9
$855.9
$418.4
$181.6
$115.6
$6,127.4
Tennessee
$6,774.7
$1,179.5
$672.4
$130.8
$6.4
$8,763.7
Texas
$30,485.1
$8,309.3
$4,090.2
$1,545.4
$721.3
$45,151.3
Utah
$2,354.1
$948.4
$701.5
$267.0
$83.4
$4,354.4
Virginia
$5,588.6
$1,341.3
$847.5
$216.4
$141.2
$8,135.2
Washington
$7,262.5
$1,763.0
$1,636.8
$728.6
$338.4
$11,729.4
Wisconsin
$5,324.9
$1,601.5
$1,037.9
$528.8
$76.1
$8,569.2
Partial Participation
States*
$17,532.5
$4,360.1
$3,025.7
$1,534.8
$407.2
$26,860.3
Subtotal
$309,624.2
$78,047.8
$46,941.6
$21,536.5
$7,403.6
$463,553.8
American Samoa
$210.4
$49.7
$23.0
$18.1
$1.2
$302.5
Guam
$97.4
$44.3
$70.5
$27.5
$30.3
$270.0
Northern Mariana Is.
$112.1
$41.7
$26.7
$14.2
$3.6
$198.4
Virgin Islands
$154.5
$12.1
$57.0
$8.4
$3.1
$235.1
Subtotal
$574.4
$147.8
$177.2
$68.1
$38.2
$1,005.9
Total
$310,198.6
$78,195.6
$47,118.9
$21,604.7
$7,441.9
$464,559.6
* The needs for partial participation states (discussed in Appendix A) are presented cumulatively and not by state. The list of 14
partial participation states is shown in Exhibit 2.4.
37
Exhibit 2.2: State 20-year Need Reported by System Size (in millions of January 2015 dollars)
State
Large
Medium
Small
NPNCWSs
Total
Alabama
$2,813.8
$8,062.1
$381.2
$4.8
$11,261.9
Arizona
$5,039.5
$3,035.1
$1,032.5
$23.4
$9,130.5
Arkansas
$977.3
$5,219.9
$1,170.4
$9.2
$7,376.8
California
$32,942.7
$13,520.4
$4,441.8
$128.3
$51,033.2
Colorado
$3,874.7
$4,850.0
$1,462.3
$1.6
$10,188.7
Connecticut
$1,807.2
$1,412.2
$763.5
$34.8
$4,017.7
District of Columbia
$1,741.9
$0.0
$0.0
$0.0
$1,741.9
Florida
$11,399.3
$8,239.0
$2,086.5
$161.6
$21,886.4
Georgia
$4,400.6
$5,969.9
$2,072.2
$17.4
$12,460.1
Illinois
$5,767.1
$11,688.8
$3,315.1
$139.4
$20,910.4
Indiana
$1,794.1
$4,227.5
$1,275.4
$223.2
$7,520.2
Iowa
$448.8
$5,559.5
$1,820.8
$23.3
$7,852.4
Kansas
$1,285.7
$2,387.6
$1,645.0
$4.4
$5,322.6
Kentucky
$1,717.1
$6,040.2
$473.3
$1.4
$8,232.0
Louisiana
$1,251.8
$4,508.9
$1,551.0
$18.9
$7,330.6
Maine
$158.9
$566.6
$579.2
$43.6
$1,348.3
Maryland
$7,388.0
$1,116.9
$700.9
$124.3
$9,330.1
Massachusetts
$2,900.2
$8,758.8
$543.8
$41.6
$12,244.4
Michigan
$4,882.4
$5,509.9
$2,056.3
$597.6
$13,046.2
Minnesota
$1,110.0
$4,322.9
$1,735.5
$339.5
$7,507.9
Mississippi
$245.8
$2,500.5
$2,064.8
$12.1
$4,823.2
Missouri
$2,334.4
$4,141.9
$2,393.0
$49.5
$8,918.9
Nevada
$3,964.0
$1,010.2
$324.5
$18.0
$5,316.8
New Jersey
$3,612.9
$3,968.8
$743.9
$257.6
$8,583.1
New York
$11,720.7
$6,506.6
$4,378.9
$159.7
$22,765.9
North Carolina
$4,772.4
$9,302.6
$2,179.4
$467.9
$16,722.2
Ohio
$5,031.9
$6,102.9
$1,913.4
$357.1
$13,405.4
Oklahoma
$1,296.3
$3,669.5
$1,864.9
$28.2
$6,858.9
Oregon
$1,761.0
$3,026.6
$1,392.4
$70.3
$6,250.4
Pennsylvania
$6,809.5
$6,378.1
$3,227.4
$356.6
$16,771.6
Puerto Rico
$1,173.7
$1,816.9
$712.0
$1.6
$3,704.2
South Carolina
$1,643.7
$3,779.1
$684.1
$20.5
$6,127.4
Tennessee
$2,048.3
$6,131.3
$547.7
$36.4
$8,763.7
Texas
$19,374.0
$18,850.6
$6,866.5
$60.3
$45,151.3
Utah
$1,269.2
$2,403.6
$665.2
$16.4
$4,354.4
Virginia
$3,114.0
$3,360.0
$1,545.1
$116.0
$8,135.2
Washington
$2,969.3
$5,593.8
$3,019.4
$146.9
$11,729.4
Wisconsin
$2,393.7
$3,881.4
$1,682.3
$611.8
$8,569.2
Partial Participation States*
$4,856.3
$12,682.9
$8,898.3
$422.9
$26,860.3
Subtotal
$174,092.1
$210,103.5
$74,210.0
$5,148.2
$463,553.8
American Samoa
$0.0
$250.3
$52.2
$0.0
$302.5
Guam
$270.0
$0.0
$0.0
$0.0
$270.0
Northern Mariana Is.
$0.0
$113.3
$85.1
$0.0
$198.4
Virgin Islands
$0.0
$162.9
$72.2
$0.0
$235.1
Subtotal
$270.0
$526.5
$209.4
$0.0
$1,005.9
Total
$174,362.0
$210,630.0
$74,419.4
$5,148.2
$464,559.6
38
Exhibit 2.3: Overview of 20-year Need by State (including District of Columbia, Puerto Rico and U.S.
Territories)
Note: Map does not include needs for American Indian and Alaska Native Village water systems.
* The list of the 14 partial participation states can be found in Exhibit 2.4.
The needs for American Samoa, Guam, the Commonwealth of the Northern Mariana Islands, and the U.S. Virgin Islands are less
than $1 billion each.
39
States that received the minimum DWSRF allotment of 1 percent in the most recent allocation were given the
option of surveying only the large systems in their states and not collecting data for medium systems. (Small
system data were collected by EPA in the 2007 Assessment.) This option was provided to reduce the burden on
these states and allow for resources to be focused on the large systems. Of the 19 states (including the District of
Columbia and Puerto Rico) that received the minimum allocation based on the 2011 Assessment findings, 14
chose this “partial participation” option. For these states, the medium system need was estimated based on data
from fully surveyed states. Because this method does not meet the Assessment’s stringent data quality objectives
at the state level, the needs of these states contribute to the estimate of the total national need but are not reported
individually by state. Exhibit 2.4 shows the large system need and small system need (extrapolated from the 2007
Assessment) estimated by state, and the total medium system need for the partial participation states.
Exhibit 2.4: State 20-year Need Reported for Partial Participation States (in millions of January 2015
dollars)
State
Medium*
Small
NPNCWSs
Alaska
$436.4
$77.3
Delaware
$354.3
$4.1
Hawaii
$192.0
$1.3
Idaho
$989.7
$47.8
Montana
$946.5
$64.1
Nebraska
$992.0
$20.2
New Hampshire
$864.4
$78.4
New Mexico
$775.8
$19.5
North Dakota
$500.0
$6.7
Rhode Island
$97.4
$20.4
South Dakota
$589.8
$6.4
Vermont
$642.7
$0.2
West Virginia
$1,074.8
$61.0
Wyoming
$442.6
$15.5
Total
$12,682.9
$8,898.3
$422.9
* The medium community water system need was estimated cumulatively based on data from fully surveyed states.
The small system need is based on the 2007 Assessment findings adjusted to 2015 dollars, an updated inventory, and updated
cost models.
The not-for-profit noncommunity water system need is based on the 1999 Assessment findings adjusted to 2015 dollars.
More of the need of the partial participation states is for their small and medium systems than their large systems,
which reflects that these states generally have few systems serving more than 100,000 people. This finding also
applies to some full-participation states with few large systems relative to the number of medium and small
systems or no very large systems that serve over 1 million people.
Needs of Water Systems in U.S. Territories
Under SDWA and through appropriations, 1.5 percent of DWSRF monies are allocated to the U.S. territories
(American Samoa, the Commonwealth of the Northern Mariana Islands, Guam, and the U.S. Virgin Islands) to
be used as grants for water systems. To assess the needs of water systems in U.S. territories, EPA collected data
from all medium and large systems in the 2015 Assessment and used the findings from the 2007 Assessment to
40
estimate the needs of small systems. Exhibit 2.5 shows the 20-year need reported for each of the U.S. territories
in millions of January 2015 dollars.
Exhibit 2.5: 20-year Need Reported by U.S. Territories (in millions of January 2015 dollars)
Territory
Total Need
American Samoa
$302.5
Guam
$270.0
Commonwealth of the
Northern Mariana Islands
$198.4
U.S. Virgin Islands
$235.1
The Assessments have consistently demonstrated that water systems in the territories face unique challenges in
providing safe drinking water to their customers. As made clear in their survey submissions and supporting
documentation, while drinking water issues can vary from island to island, the overall challenges for all the U.S.
territories include rapidly deteriorating infrastructure, a seasonal transient customer base, limited source water
options, and ground water contamination.
Changes in State-Specific Need through Assessment Cycles
The state-specific results of the Assessments show that states’ needs change, some more significantly than others,
during the four-year intervals between Assessments (see Exhibit 2.6). State-specific changes from one
Assessment to the next as well as needs relative to each other can be attributed to two primary factors:
Changes in Projects Planned, Initiated, and Completed. Congress specified that the DWINSA be repeated at
4-year intervals to capture changes in system infrastructure needs. Changes in the reported needs of individual
systems from one Assessment period to the next can have a significant effect on the overall state need. For
example, in one Assessment a large system that has identified a project with very substantial costs might cause
that state’s need to increase due to the large project. However, if construction of the project begins prior to the
next Assessment cycle, those needs would not be included in the next Assessment. If all other needs in that state
were to remain constant, the state’s need would be lower for the next Assessment. In addition, conditions such as
drought or other momentous events that affect water systems within a state may change significantly over a four-
year period and have an impact on that state’s need from projects planned to mitigate or address the event.
Changes in National and State Assessment Approaches. State-specific needs have also been affected by
changes in the Assessment methodology since the first initial effort in 1995. The Assessment’s “bottom-up”
approach of receiving documented needs on a project-by-project basis for each system has remained essentially
unchanged, and the general statistical approach of stratified random sampling of water systems has also
remained unchanged, with the exception of applying the panel approach to medium system samples in full
participation states in this 2015 DWINSA (as described in Appendix B).
However, since the first effort in 1995, changes to the Assessment’s general policies and procedures have
affected individual states’ infrastructure needs, both in terms of their total state need and their relative share of
the total national need. Changes have been made to the types of parties responsible for data collection, the type
of documentation required to support acceptance of an identified need, and policies and approaches
implemented to further ensure both complete and high-quality data collection. The evolving methodology over
the last 20 years, from 1995 to 2015, is further detailed in Appendix A and Appendix B.
41
Exhibit 2.6: Historic State Need Reported for Each Assessment (20-year need in millions of January 2015
dollars)
State
1995
1999
2003
2007
2011
2015
Alabama
$3,040
$1,796
$2,559
$5,188
$8,869
$11,262
Alaska
$1,413
$973
$1,033
$1,028
*
*
Arizona
$2,480
$2,696
$13,819
$9,378
$8,301
$9,131
Arkansas
$3,709
$2,550
$5,362
$6,680
$6,804
$7,377
California
$34,469
$29,067
$42,233
$49,412
$49,663
$51,033
Colorado
$3,571
$4,206
$8,067
$8,099
$7,948
$10,189
Connecticut
$2,486
$1,673
$990
$1,764
$3,992
$4,018
Delaware
$681
$505
$365
*
*
*
District of Columbia
$241
$688
$226
$1,106
$1,793
$1,742
Florida
$7,943
$6,190
$22,791
$16,227
$18,376
$21,886
Georgia
$6,036
$3,999
$13,664
$11,311
$10,340
$12,460
Hawaii
$789
$244
$1,231
*
*
*
Idaho
$1,081
$857
$1,102
*
*
*
Illinois
$9,801
$10,221
$20,451
$19,004
$21,181
$20,910
Indiana
$3,068
$2,815
$6,109
$7,523
$7,304
$7,520
Iowa
$4,133
$4,731
$5,309
$7,736
$6,616
$7,852
Kansas
$3,621
$2,735
$2,926
$5,100
$4,680
$5,323
Kentucky
$4,075
$2,941
$4,256
$6,300
$6,949
$8,232
Louisiana
$3,579
$2,115
$6,223
$8,732
$5,938
$7,331
Maine
$1,586
$829
$1,260
*
$1,316
$1,348
Maryland
$2,354
$2,777
$6,005
$6,889
$7,713
$9,330
Massachusetts
$10,892
$9,767
$12,963
$8,593
$8,592
$12,244
Michigan
$8,129
$11,282
$17,139
$14,987
$15,412
$13,046
Minnesota
$4,466
$5,151
$8,274
$7,578
$8,214
$7,508
Mississippi
$2,888
$2,262
$2,492
$4,104
$4,113
$4,823
Missouri
$3,442
$3,623
$9,028
$8,967
$9,462
$8,919
Montana
$1,214
$1,449
$1,196
*
*
*
Nebraska
$1,746
$1,383
$2,052
$2,248
*
*
Nevada
$962
$1,001
$1,382
$3,406
$6,238
$5,317
New Hampshire
$1,314
$830
$902
*
*
*
New Jersey
$6,620
$6,081
$10,479
$10,075
$8,830
$8,583
New Mexico
$1,910
$1,732
$1,397
*
*
*
New York
$18,472
$21,864
$22,445
$34,342
$24,591
$22,766
North Carolina
$4,972
$4,499
$16,638
$12,725
$11,208
$16,722
North Dakota
$1,075
$814
$920
*
*
*
Ohio
$8,989
$8,242
$14,674
$15,944
$13,602
$13,405
Oklahoma
$3,722
$3,890
$7,280
$5,204
$7,245
$6,859
Oregon
$3,936
$4,503
$6,467
$3,525
$6,207
$6,250
Pennsylvania
$8,713
$8,739
$16,653
$14,400
$15,873
$16,772
42
State
1995
1999
2003
2007
2011
2015
Puerto Rico
$4,130
$3,277
$3,453
$3,211
$3,583
$3,704
Rhode Island
$1,203
$959
$610
*
*
*
South Carolina
$2,676
$1,364
$1,887
$2,061
$0
$6,127
South Dakota
$1,042
$731
$1,500
*
*
*
Tennessee
$3,428
$2,344
$4,198
$4,489
$3,003
$8,764
Texas
$22,653
$21,718
$42,685
$33,068
$37,813
$45,151
Utah
$1,915
$854
$1,071
*
$4,157
$4,354
Vermont
$841
$510
$598
*
*
*
Virginia
$5,393
$3,416
$4,341
$7,671
$7,493
$8,135
Washington
$7,385
$6,561
$10,109
$12,346
$10,621
$11,729
West Virginia
$1,997
$1,695
$1,306
*
*
*
Wisconsin
$3,421
$5,149
$8,998
$7,828
$7,967
$8,569
Wyoming
$716
$735
$452
*
*
*
Partial Participation States*
$21,757
$26,735
$26,860
Subtotal
$250,416
$231,032
$399,571
$410,005
$418,745
$463,554
American Samoa
$41
$60
$49
$117
$91
$302
Guam
$195
$191
$423
$334
$263
$270
North Mariana Is.
$64
$124
$300
$366
$198
$198
Virgin Islands
$409
$269
$273
$321
$195
$235
Subtotal
$710
$644
$1,045
$1,138
$747
$1,006
Total
$251,126
$231,676
$400,616
$411,144
$419,492
$464,560
Note: This table represents needs of state systems only; needs of American Indian and Alaska Native Village water systems are
reported in Chapter 3. In addition, this table does not include needs associated with proposed or recently promulgated regulations.
* For the 2007, 2011, and 2015 Assessments, the need for partial participation states (discussed in Appendix A) that opted out of the
medium system portion of the survey is presented cumulatively and not by state.
43
Chapter 3: American Indian and Alaska
Native Village Need
American Indian and Alaska Native Village-Specific Needs
The combined American Indian and Alaska Native Village water system need estimated for the 2015 Assessment
is $3.8 billion in capital improvements over the next 20 years. This need includes drinking water infrastructure to
increase access to safe drinking water through compliance with EPA’s drinking water regulations, as well as
connection of homes without piped water to existing public water systems. Exhibit 3.1 presents the American
Indian and Alaska Native Village water system need by EPA regional offices and by project category.
The American Indian and Alaska Native Village water system needs are extrapolated from the findings of the
2011 Assessment, which were based on statistically designed surveys of American Indian and Alaska Native
Village water systems. These were the first surveys of these systems since 1999, and they incorporated the changes
to EPA’s approach and policies for estimating infrastructure needs of non-tribal systems between 1999 and 2011.
Extrapolation from the 2011 findings was based on the Construction Cost Index (CCI) compiled by McGraw Hill
Construction and application of current cost models.
Exhibit 3.1: 20-year Need for American Indian by EPA Region and Alaska Native Village Systems (in
millions of January 2015 dollars)
EPA Region
Transmission and
Distribution
Source
Treatment
Storage
Other
Total Need
Region 1
$3.6
$0.9
$1.1
$1.0
$0.2
$6.8
Region 2
$20.5
$1.8
$2.1
$3.0
$1.2
$28.6
Region 3
*
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
Region 4
$29.4
$6.3
$9.7
$8.3
$1.6
$55.4
Region 5
$123.6
$18.6
$28.1
$33.4
$6.7
$210.4
Region 6
$99.1
$15.0
$29.7
$32.4
$6.4
$182.6
Region 7
$21.8
$2.6
$4.0
$5.6
$1.2
$35.3
Region 8
$380.7
$27.9
$76.5
$79.8
$6.7
$571.7
Region 9
$1,323.6
$91.4
$178.2
$186.0
$38.0
$1,817.1
Region 10
$129.6
$19.7
$35.8
$32.9
$6.5
$224.5
Subtotal - American
Indian Systems
$2,132.0
$184.3
$365.3
$382.4
$68.4
$3,132.4
Alaska Native Village
Systems
$303.8
$43.7
$190.6
$118.9
$5.5
$662.5
Total
$2,435.8
$228.1
$555.8
$501.3
$73.9
$3,795.0
Note: Numbers may not total due to rounding.
*
There are no American Indian water systems in EPA Region 3.
Navajo water systems are located in EPA Regions 6, 8, and 9, but for purposes of this report, all Navajo water system needs are
reported in EPA Region 9.
Needs for Alaska Native Village water systems are not included in the EPA Region 10 total.
44
American Indian Needs
The total 20-year need for American Indian water systems is estimated to be $3.1 billion. Exhibit 3.2 shows the
total American Indian water system need by project category. The American Indian need includes substantial
infrastructure to increase access to safe drinking water through connection of homes without water to existing
public water systems. As would be expected for these systems, transmission and distribution is the largest project
category, representing 68 percent of the total need. This high percentage reflects the significant infrastructure and
logistical challenges associated with American Indian water systems, which serve widely dispersed populations
in remote locations.
Exhibit 3.2: Total 20-year Need by Project Category for American Indian Water Systems (in millions of
January 2015 dollars)
Note: Numbers may not total due to rounding.
Alaska Native Village Needs
The 2015 total 20-year need for Alaska Native Village water systems is estimated to be $0.7 billion. Exhibit 3.3
shows the total Alaska Native Village water system need by project category. The need for Alaska Native Village
water systems differs from more typical CWSs in that costs for piping make up less than half the need, with
storage and treatment comprising a greater percentage of the total. These smaller communities with homes in
close proximity typically have lower relative costs for piping. They face higher treatment and storage costs in
many locations to accommodate seasonal source water availability. Piping, treatment, and storage costs for same-
size projects are higher than typical for the states because of the remote locations or arctic conditions of Alaska
Native Villages.
45
Exhibit 3.3: Total 20-year Need by Project Category for Alaska Native Village Water Systems (in millions
of January 2015 dollars)
Note: Numbers may not total due to rounding.
Pipe used for water main replacement in Houston,
TX.
46
Chapter 4 Conclusions
EPA, along with state, tribal and local stakeholders strive to ensure each Assessment accurately reflects the true
DWSRF-eligible need of small, medium, and large public water systems throughout the country, including those
serving American Indian and Alaska Native communities and the U.S. territories. EPA and its partners also strive
to capture emerging drinking water industry challenges and assess those impacts on the total national need. Survey
response data and feedback from the DWINSA workgroup identify trends and data gaps, allowing EPA to improve
and refine the design of future surveys to more fully capture and assess system needs.
The DWINSA is a significant endeavor that captures responses from more than 2,500 large and medium water
systems (as well as small systems and systems in American Indian and Alaska Native Villages in previous
Assessments). A response rate of nearly 100 percent from systems surveyed, as well as other information,
demonstrates a current 20-year need for funding of DWSRF-eligible projects of over $470 billion. This
quadrennial nation-wide survey is uniquely positioned to capture changing industry conditions in the future.
In addition to its fundamental purpose as a data collection instrument, the Assessment also strives to serve as a
useful tool for utilities by promoting asset management, including the development of a record of system
infrastructure assets and their condition. EPA has received feedback from partners that the approach to the survey
has been successful in contributing to enhanced asset management in the drinking water sector. EPA looks
forward to working with partners to build on this success to further improve estimates of overall infrastructure
investment need for projects that can be funded through the DWSRF and to continue to enhance infrastructure
asset management for the sector in the future.
47
Appendix A - Survey Methods
The 1996 Safe Drinking Water Act (SDWA) Amendments direct the U.S. Environmental Protection Agency
(EPA) to assess the needs of water systems and to use the results of the quadrennial Assessment to allocate
Drinking Water State Revolving Fund (DWSRF) monies. The DWSRF monies are allocated based on each state’s
share of the total state need with a minimum of 1 percent of the state allotment guaranteed to each state, Puerto
Rico, and the District of Columbia. The results of the Assessment are also used to allocate the percentage (recently
1.5 percent) of the DWSRF appropriation designated for the U.S. territories (the U.S. Virgin Islands, Guam,
American Samoa, and the Commonwealth of the Northern Mariana Islands). Further, the results of the Assessment
are used, in part, to allocate the DWSRF appropriation (recently 2 percent) designated for the American Indian
and Alaska Native Villages to nine EPA regional offices for grants to these water systems (EPA Region 3 does
not have any federally recognized tribes). The Drinking Water Infrastructure Needs Survey and Assessment
(DWINSA or Assessment) estimates the need for both community water systems (CWSs) and not-for-profit
noncommunity water systems (NPNCWSs).
The 20-year period captured by the 2015 Assessment is from January 1, 2015, through December 31, 2034. The
Assessment is based on 2,562 responses to a survey of large and medium water systems in states, Puerto Rico,
the District of Columbia, and U.S. territories. It also included an estimate of small system needs using data from
the 2007 Assessment for small water systems, and an estimate of the needs of NPNCWSs using data from the
1999 Assessment. The 2015 Assessment also included an estimate of the needs of American Indian and Alaska
Native Village water systems based on data from the 2011 Assessment.
The Assessment was developed in consultation with a workgroup consisting of state and EPA regional
coordinators. The workgroup met several times by conference call and in person and reached a final consensus
on the Assessment’s policies and processes. Except where noted, the basic statistical and survey methodologies
of the 2015 Assessment are nearly identical to those used in previous Assessments. The 2015 Assessment used
the same survey method for the large systems as for past Assessments; however, a slightly modified approach
was used for medium systems, which is described in more detail later in this Appendix. The questionnaire used
in the 2015 Assessment was essentially the same as the 2003, 2007, and 2011 questionnaires.
In compliance with the Paperwork Reduction Act (44U.S.C. 3501 et seq.), the survey design and instrument were
reviewed and approved by the Office of Management and Budget (OMB). The Information Collection Request
for the survey can be accessed in the Federal Register (80 FR 13538; March 16, 2015).
Assessing the Needs of Water Systems in States and U.S. Territories
Statistical Frame
The frame is the list of all members (sampling units) of a population from which a sample will be drawn for a
survey. For this Assessment, the frame consisted of all medium and large CWSs in each state, Puerto Rico, the
District of Columbia, and the U.S. territories. As discussed below, this Assessment used the result of the 2007
Assessment for small CWSs and excluded small systems from this survey’s frame. American Indian and Alaska
Native Village systems were not included in the 2015 survey; as discussed below, the results from the 2011
Assessment were used to estimate needs for these systems.
To ensure that the survey accounted for all medium and large CWSs in the nation, the universe of water systems
was obtained from the federal Safe Drinking Water Information System (SDWIS-FED). SDWIS-FED is EPA’s
48
centralized database of public water systems. It includes the inventory of all public water systems and provides
information on the size of the population served by each system and each system’s water source (ground water,
surface water, or both).
Each state was asked to review the frame and verify or correct information on each system’s source water and
population served. EPA used this updated information to create a complete list of medium and large CWSs. A
sample of systems was then selected from this updated frame.
Modified Panel Approach and the Stratified Sample
The inventory of systems from which the sample was drawn included 9,249 CWSs, including medium and large
systems in states that fully participated and large systems in the partial participation states (see “Conducting the
Survey of Medium Systems” below for more information about full and partial participation states). The inventory
of 9,249 systems does not include the 597 medium systems in partial participations states, nor small systems.
Because it would be prohibitively expensive to collect information from every system, EPA collects information
about the infrastructure investment needs from a sample of these systems. The 2015 Assessment was designed to
achieve a desired level of precision for state-level estimates of total infrastructure needs for medium-size systems
serving between 3,301 and 100,000 people. EPA used a modified panel approach that includes a census of large
systems and a statistical sample of medium systems to estimate total capital needs. This statistical approach
minimizes burden while achieving the desired level of precision.
A panel (or longitudinal) approach tracks a cross-section of observations over time and is common in econometric
studies of individuals, households, firms, and utilities. It provides advantages in terms of survey/assessment costs
and reduced burden, while allowing for trend analyses and providing probability samples of systems on which
estimates of each state’s needs can be based. For the 2015 Assessment, rather than selecting a completely new
sample of systems, EPA reassessed the needs of most of the systems that participated in the 2011 Assessment.
This approach reduced the amount of time required for systems to prepare and for states to review the responses
from systems resurveyed in 2015. While the panel approach reduces the burden of the survey, it does have a
potential drawback. If a completely new sample were selected for each Assessment, the sampling error would be
a random component that changed from survey to survey. With a panel approach, this error becomes systematic.
To partially alleviate this potential source of bias, EPA modified the panel approach by replacing 25 percent of
the sample of systems serving 3,301 to 100,000 people. This approach ensured that EPA could continue to meet
the precision targets for each state.
To meet the state-level precision targets, EPA started with the sample of systems used in the 2011 Assessment.
EPA adjusted the sample size required to accommodate changes in the inventory of systems, including the
addition of new systems, the removal of systems that are no longer active, and the movement of systems among
strata (groups). As in the 2011 Assessment, EPA first determined the total sample size needed for each state to
meet the target level of precision. EPA then allocated the sample among strata in order to maximize the efficiency
of the design.
To determine aggregated needs, water systems were divided into strata by size (population served) and by source
(surface water or ground water). Exhibit A.1 shows the population and source water strata for the state survey.
49
Exhibit A.1: Stratification of the State Community Water System Survey
* In some states, systems serving 10,001 50,000 can be considered one stratum and precision targets can be met. The most efficient
sample is drawn for each state.
EPA designed the sample to estimate the need for each state, Puerto Rico, the District of Columbia, and the U.S.
territories with a precision target of ± 10 percent with 95 percent confidence. To meet this target, all large systems
(serving more than 100,000 people) were surveyed. These systems have the largest infrastructure needs and have
the staff to respond efficiently to the 2015 Assessment. EPA also surveyed a stratified random sample of medium
systems (serving 3,301 to 100,000 people) in each fully participating state and, because there are so few, all of
the medium and large systems in the U.S. territories. This methodology further reduces burden and still achieves
the Assessment data quality objectives.
The population served by each system includes a system’s retail customers and any consecutive populations
(people that are served by other water systems that purchase water from the system). Consecutive populations are
included in the system population because critical infrastructure of the wholesale system would need to be sized
to accommodate the demand of the population directly served by the consecutive system.
Systems are categorized as surface water if they have at least one source that is surface water or ground water
under the direct influence of surface water (GWUDI). Systems are categorized as ground water if they do not
have a surface water or GWUDI source. The ground water category includes ground water systems and systems
that do not have a source of their own and purchase finished water from another system (regardless of whether
the purchased water comes from a surface water or ground water source). The decision to include these
consecutive systems that purchase water in the ground water systems category was based on the 1995
Assessment’s findings that, in general, the needs of purchased water systems more closely resemble those of
ground water systems than those of surface water systems that need to provide treatment to comply with the
Surface Water Treatment Rules.
System Weights
For the large and medium systems surveyed, the 2015 Assessment assigned weights to the findings from each
surveyed water system to determine total state needs. Because all large systems are included in the survey, each
large system has a weight of one. The number of medium water systems selected from each stratum was
determined by the total number of systems in that stratum (see Exhibit A.2), the percentage of that state’s need
represented by that stratum in the most recent Assessment, and the relative variance of the need within that stratum
3,301-10,000
10,001-25,000
or 10,001-
50,000*
25,001-50,000
50,001-100,000
Medium
Systems
>100,000
Large
Systems
Ground Water
Surface Water
Population
State-specific samples for participating states
Sampled with certainty - All systems receive questionnaire
50
in the most recent Assessment. The sample is allocated among the strata in a manner that lets the survey achieve
the desired level of precision with the smallest sample size for each state. For medium systems, the sampled
systems were selected by stratum and assigned an initial weight equal to the total number of systems in that
stratum divided by the number of systems in that stratum’s sample. For medium systems, a final weight was
recalculated for each stratum with adjustments for non-response as discussed in more detail below.
Each fully surveyed state’s need was determined by summing the cost of each project for each system, and then
multiplying each system’s need by the system’s final weight. Relatively small changes in need in a medium
system with a high weight can result in a significant impact on a state’s need. The need for each project is
multiplied by the system’s statistical weight. The higher the weight, the higher the impact.
Final Sample
During data collection, new information provided by responding systems changes the sample such that the final
sample varies from the initial. In addition, some systems do not respond to the survey request and this also changes
the final sample. Exhibit A.2 shows the initial number of systems in each sampling stratum, the sample target
size, non-responding systems, systems removed from the sample, migrating systems, and the final sample size.
Details of the adjustments to the sample are provided below.
Exhibit A.2: Number of Systems in Frame and Sample by Source Type and Population Served
Source
Type
Size
Category
Population
Served
Number
of
Systems
in the
Frame
Sample
Required to
Meet
Precision
Targets
Non-
Response
Net Strata
Migration
Systems
Removed
from
Sample
Final
Sample
Medium
3,301-10,000
3,622
519
1
14
11
521
Ground
10,001-
50,000*
2,490
588
1
15
7
595
Water
50,001-
100,000
396
153
0
0
1
152
Large
More than
100,000
269
269
0
-31
5
233
Medium
3,301-10,000
843
165
0
8
2
171
Surface
10,001-
50,000*
1,043
307
5
18
1
319
Water
50,001-
100,000
298
160
1
1
0
160
Large
More than
100,000
439
439
0
-25
3
411
Total
9,400
2,600
8
0
30
2,562
*
In some states, systems serving 10,001-50,000 people are further divided into two population categories: systems serving 10,001-
25,000 people and systems serving 25,001-50,000 people.
51
For the 2015 Assessment, eight of the 2,600 selected systems did not respond to the survey, for an overall response
rate of 99.7 percent. For systems that did respond, there were instances where systems reported a different water
source or a significantly different population compared to the original information, causing systems to migrate
across sampling strata. In addition, based on their response to the survey, 30 systems were removed from the final
sample for the following reasons:
Nineteen systems reported that the population they serve is less than 3,301. (The sample includes only
systems serving more than 3,300 people.)
Four systems in partial participation states reported a population of less than 100,000. (The sample does
not include systems serving 3,301 to 100,000 people in states that do not fully participate in the survey.)
Six systems reported needs with another water system. As such, their needs are accounted for elsewhere.
As an example, New York City is served by three CWSs but they choose to submit one combined
questionnaire for the three systems.
One system was not a CWS and should not have been included in the frame.
The final sample that was used to estimate the national need for large and medium systems was 2,562 as shown
in Exhibit A.2.
Conducting the Survey of Large Systems
Systems serving more than 100,000 people, including retail and consecutive populations, were considered large
systems for the 2015 Assessment. All large systems are selected for the survey because of the unique nature of
systems in this size category and because they represent a large portion of the nation’s need. The population cut-
off was the same threshold as used in the 2007 and 2011 Assessments. In the 1995, 1999, and 2003 Assessments,
the large system category was defined as systems serving populations of more than 40,000 or 50,000.
EPA provided the survey instruments to each state for each large system in the state. States worked with each
system to complete the questionnaire and collect and prepare documentation for the projects listed on the
questionnaire. The state coordinators reviewed the questionnaires to ensure that the systems included all their
needs, the information entered on the questionnaire was correct, the projects were eligible for DWSRF funding,
and the projects were adequately documented. During their reviews, states often contacted systems to obtain
additional information. The states then submitted the questionnaire and all documentation to EPA for a final
review.
All of the 708 large systems that received a survey for the 2015 Assessment responded to the survey. Based on
the survey responses and migration of systems between strata, the final sample include 644 systems that serve
more than 100,000 people. The number of large systems in the final sample for each state, Puerto Rico, the District
of Columbia, and the U.S. territories is shown in Exhibit A.3.
52
Exhibit A.3: Estimated Number of Medium and Large Systems in the Inventory and the Final Sample by
State
State or Territory
Estimated Inventory of Systems
Final Sample
Serving
3,301-
100,000
People
Serving
More than
100,000
People
Total # of
Medium
and Large
Systems
Serving
3,301-
100,000
People
Serving
More than
100,000
People
Total # of
Medium
and Large
Systems
Alabama
366
23
389
120
23
143
Alaska
0
1
1
0
1
1
Arizona
124
10
134
22
10
32
Arkansas
196
5
201
85
5
90
California
552
135
687
37
135
172
Colorado
151
15
166
35
15
50
Connecticut
50
10
60
23
10
33
Delaware
0
3
3
0
3
3
District of Columbia
0
1
1
0
1
1
Florida
304
56
360
33
56
89
Georgia
228
22
250
41
22
63
Hawaii
0
2
2
0
2
2
Idaho
0
1
1
0
1
1
Illinois
469
21
490
33
21
54
Indiana
214
9
223
65
9
74
Iowa
147
2
149
54
2
56
Kansas
109
6
115
75
6
81
Kentucky
235
6
241
83
6
89
Louisiana
252
7
259
74
7
81
Maine
35
1
36
22
1
23
Maryland
55
5
60
25
5
30
Massachusetts
249
9
258
53
9
62
Michigan
289
15
304
37
15
52
Minnesota
171
4
175
100
4
104
Mississippi
204
1
205
87
1
88
Missouri
224
7
231
87
7
94
Montana
0
1
1
0
1
1
Nebraska
0
2
2
0
2
2
Nevada
35
5
40
14
5
19
New Hampshire
0
1
1
0
1
1
New Jersey
232
16
248
39
16
55
New Mexico
0
1
1
0
1
1
New York
318
24
342
24
24
48
North Carolina
320
22
342
72
22
94
North Dakota
0
1
1
0
1
1
Ohio
314
17
331
50
17
67
53
Estimated Inventory of Systems
Final Sample
State or Territory
Serving
3,301-
Serving
More than
Total # of
Medium
Serving
3,301-
Serving
More than
Total # of
Medium
100,000
100,000
and Large
100,000
100,000
and Large
People
People
Systems
People
People
Systems
Oklahoma
172
5
177
63
5
68
Oregon
109
7
116
36
7
43
Pennsylvania
302
20
322
36
20
56
Puerto Rico
102
5
107
49
5
54
Rhode Island
0
2
2
0
2
2
South Carolina
159
9
168
21
9
30
South Dakota
0
2
2
0
2
2
Tennessee
253
12
265
116
12
128
Texas
1,021
65
1,086
74
65
139
Utah
106
11
117
21
11
32
Virginia
147
19
166
32
19
51
Washington
211
11
222
44
11
55
West Virginia
0
1
1
0
1
1
Wisconsin
176
7
183
32
7
39
Subtotal
8,601
643
9,244
1,914
643
2557
American Samoa
1
0
1
1
0
1
Guam
0
1
1
0
1
1
Northern Mariana Is.
1
0
1
1
0
1
Virgin Islands
2
0
2
2
0
2
Subtotal
4
1
5
4
1
5
Total
8,605
644
9,249
1,918
644
2,562
Conducting the Survey of Medium Systems
Medium systems, as defined for the 2007, 2011, and 2015 Assessments, serve 3,301 to 100,000 people. Exhibit
A.3
shows the number of medium systems in the final sample by state and an estimate of the total number of
medium systems in each state. (States with zeros in the medium system sample column opted not to collect data
for these systems.)
For the 2015 Assessment, states that received the minimum 1 percent DWSRF allotment in the 2011 Assessment
were given the option of not participating in data collection for medium systems. This option was provided to
reduce the burden of the Assessment on states whose allocation is unlikely to be affected by the findings of the
survey. Of the minimum allocation states, 14 chose not to participate in this portion of the survey. The medium
system need for states that chose this option was estimated using data from participating states and the inventory
of medium systems in the partial participation states. Use of this method allowed EPA to meet its precision target
for each full participation state as well as at the national level. It does not meet precision targets for the partial
participation states and therefore their needs are not reported individually by state.
54
For states that participated in the medium system portion of the survey, the data collection process was similar to
that of large systems. Survey instruments were provided to the states for the medium systems in each state’s
sample. States worked with the systems to complete the questionnaire and collect and prepare documentation for
the projects listed on the questionnaire.
Once the need for medium systems in the fully surveyed states was calculated, EPA used it to estimate the need
for the partial participation states. An average need per system was calculated by stratum using data from the fully
participating states and applied to the inventory of systems in the partial participation states. The inventory of
systems was verified by state personnel.
Of the 1,892 medium systems that were randomly selected and received a survey, responses were received for
1,884 systems for a response rate of 99.6 percent. Based on the survey response and migration of systems between
strata, 1,918 of the survey responses were for systems that serve 3,301 to 100,000 people. The number of medium
systems in the final sample for each state, Puerto Rico, the District of Columbia, and the U.S. territories is shown
in Exhibit A.3.
Assessing the Need for Small Systems
The infrastructure need reported for small systems serving 3,300 people or fewer is based on the findings of the
2007 Assessment when data were collected through a field survey of 600 small water systems. The 2007 small
system sample was designed to estimate the national need of small systems with a 95 percent confidence interval
of plus or minus 25 percent. Because the field survey data are believed to be highly accurate and due to resource
constraints, EPA did not survey these systems again in 2015. Instead, EPA used the projects reported for the 2007
Assessment, applied the 2015 cost models, converted all costs to 2015 dollars, calculated the average need per
small system, and multiplied this average by the number of small systems in the 2015 inventory to estimate the
2015 needs for these systems.
Assessing the Need of Not-for-Profit Noncommunity Water Systems
NPNCWSs are eligible for DWSRF funding. The 2015 estimate of need of NPNCWSs was based on the findings
of the 1999 Assessment, which surveyed a statistical sample of these systems. These findings were adjusted to
January 2015 dollars using the Construction Cost Index. The national need for NPNCWSs was allocated among
the states in proportion to the 1999 inventory of NPNCWSs in each state in a manner similar to that used for small
systems with an adjustment to 2015 dollars.
During the 1999 Assessment, EPA collected data from a national sample of 100 NPNCWSs through site visits.
Unlike the sampling design for CWSs, the NPNCWS sample was not stratified into size and source categories
because EPA lacked the empirical information on variance necessary for developing strata. The sample used for
the 1999 Assessment for NPNCWSs was designed to provide a 95 percent confidence interval that is within a
range of ± 30 percent of the estimated need.
Very little information about the needs of NPNCWS’s was available before the 1999 Assessment, including data
on the number of not-for-profit systems, the type of infrastructure needed, or their total need. The cost of collecting
this information is high because there are a large number of non-community systems and site visits would be
required to collect information about their projects and needs. Despite their large number, it was expected that
their total need would be relatively low and would have a small impact on the estimate of the national and each
state’s needs. In 1999, EPA selected a small sample of NPNCWSs to minimize the cost of collecting these data.
Therefore, the margin of error for NPNCWSs is larger than for CWSs. In addition to the high margin of error, the
55
fact that the data are more than 16 years old and that the inventory of NPNCWS’s and their total need may have
changed significantly since the 1999 Assessment diminishes the confidence in the need estimate. Due to the small
total need for NPNCWSs, this is likely to have a small effect on the estimated need nationally and of each state.
Assessing the Need of American Indian and Alaska Native Village Water Systems
The infrastructure need reported for American Indian and Alaska Native Village systems is based on the findings
of the 2011 Assessment. For the 2011 Assessment, EPA regional offices provided assistance in data collection.
Because of resource constraints, EPA did not survey these systems again in 2015. Instead, EPA used the projects
reported for the 2011 Assessment, applied the 2015 cost models, and converted all costs to 2015 dollars to estimate
the 2015 needs for these systems.
56
Appendix B - Data Collection
To determine the scope of water systems’ 20-year need, data are collected on capital improvement projects. States
and other agencies work with the surveyed systems to identify applicable projects. To be included in the U.S.
Environmental Protection Agency (EPA)’s Assessments, each project had to meet each of the following four
criteria:
The project must be for capital improvement.
The project must be eligible for Drinking Water State Revolving Fund (DWSRF) funding.
The project must be in furtherance of the public health protection goals of the Safe Drinking Water Act
(SDWA).
The project must be submitted with supporting information that documents that the three other criteria are
met.
Projects included in the Assessment generally fall into one of two categories that describe the reason for the
project:
Replacement or rehabilitation of existing infrastructure due to age or deterioration.
New or expanded infrastructure to meet an unmet need for the current population or to comply with an
existing regulatory requirement.
Projects for infrastructure generally expected to need rehabilitation or replacement in the 20-year period covered
by the Assessment were accepted with minimal documentation describing their scope and the reason for the need.
However, other types of projects required independently generated documentation that not only identified the
need but also showed clear commitment to the project by the water system’s decision-makers. Exhibit B.1
summarizes the types of projects that were included and the types that were unallowable.
57
Exhibit B.1: DWINSA Allowable and Unallowable Projects
DWINSA Allowable Projects
DWINSA Unallowable Projects
Criteria:
Eligible for DWSRF funding
Capital improvement needs
In furtherance of the public health
goals of the SDWA
Within the Assessment time frame
Adequate documentation
Project Types:
New or expanded/upgraded
infrastructure to meet the needs of
existing customers
Replacement or rehabilitation of
existing undersized or deteriorated
infrastructure
Raw water reservoir- or dam-related needs
Projects needed primarily to serve future population growth
Projects solely for fire suppression
Projects for source water protection
Non-capital needs (including studies, operation and maintenance)
Needs not related to furthering the SDWA’s public health
objectives
Acquisition of existing infrastructure
Projects not the responsibility of the water system
Projects or portions of projects started prior to January 1, 2015
Projects or portions of projects needed after December 31, 2034
For the purposes of assigning a cost to each need, the survey required that the water system either provide an
existing documented cost estimate or the information necessary for EPA to assign a cost. This information was
referred to as the “design parameter” and is discussed in more detail in this Appendix.
Survey Instrument
As with previous Assessments, the 2015 questionnaire was the survey instrument for reporting all needs. The
states were provided the survey package, which included an electronic file questionnaire for each system in the
sample, instructions for completing the questionnaire, and a list of codes used to convert the information to a
database format. For systems that completed a survey in 2011 (all large systems and the medium systems in the
panel), the state was provided an Excel version of the questionnaire with the data from the 2011 Assessment
prepopulated to facilitate updating the needs for the 2015 Assessment. Similar documents were also used by the
site visitors for recording small system needs in the 2007 Assessment, as well as for all American Indian and
Alaska Native Village water systems in the 2011 Assessment.
The instructions included information on the background and purpose of the Assessment as well as how to identify
projects that should be included in the questionnaire. Each state determined how the information would be passed
on to the water systems. In addition to infrastructure needs, the survey also requested basic information from the
water systems such as the size of the population served, the number of service connections, the production
capacity, the source water type, and the system’s ownership type. This information was compared to the
information used for the sample frame. Discrepancies in source and population were investigated to ensure
accurate information was used to draw the statistical sample.
Project Documentation
Each project listed on the questionnaire was required to have accompanying written documentation of its scope
and why it was needed. Written documentation included statements by the water system or the state as well as
more formal documents such as master plans, capital improvement plans, sanitary survey reports, and other
58
sources of project information. Whether the documentation could be written for the 2015 Assessment or had to
be pre-existing depended on project type. EPA reviewed all documentation for every project to ensure that the
project met the allowability criteria for the Assessment. See Appendix C for more information on the project
allowability policies.
Cost Estimates and Modeling
As with previous Assessments, costs assigned to projects were obtained in one of two ways. If the system had an
existing documented cost estimate that met the documentation criteria of the survey, EPA adjusted this cost to
2015 dollars and used the adjusted cost for that system’s need. This is the preferred approach for assigning a cost
to a project. If no cost estimate was available, the system was asked to provide information (design parameters)
necessary for EPA to model the cost of the project. Cost models were built from the documented cost estimates
provided by other survey respondents.
Acceptable forms of documentation for cost estimates were capital improvement plans, master plans, preliminary
engineering reports, facility plans, bid tabulations, and engineer’s estimates that were not developed for the 2015
Assessment. For each project with an associated cost, EPA needed the month and year of the cost estimate in
order to allow an adjustment of the cost to January 2015 dollars.
Systems that had cost estimates were encouraged to submit design parameters regarding size or capacity of the
needed infrastructure. For example, a storage tank is described in terms of volume in millions of gallons, treatment
plants are based on capacity in millions of gallons per day, and pipe parameters are in diameter and length. Over
70 project types were used to describe projects and link design parameters to cost. EPA used this combination of
the specific type of project, costs, and parameters as input to develop cost models. Prior to input to the cost models,
the cost estimates were normalized for both time frame and location. Cost estimates prior to January 2015 were
adjusted to January 2015 dollars using the Construction Cost Index. Regional variations in construction costs
were normalized by location using the RS Means Location Factors Index. RS Means is a subsidiary of Gordian,
which publishes an annual index used to calculate construction costs for a specific location. The factor multiplier
is expressed as a relationship to the national average of one.
The vast majority of the over 70 different types of need could be modeled by EPA. The very few project types
that could not be modeled were unique to individual systems and did not lend themselves to modeling (examples
include de-stratification of a surface water source and off-stream raw water reservoirs).
Ultimately, some projects were not assigned a cost because the system did not provide a cost estimate and project
information submitted on the survey did not include the necessary design parameters required for modeling.
Website and Database
EPA used a 2015 survey-specific website to provide an efficient method of tracking and monitoring questionnaire
responses. The website allowed controlled viewing of survey information and provided a means to submit
additional project information. State contacts and EPA had secure login access to the website. The website was a
modification of the one used successfully for the 2003, 2007, and 2011 Assessments.
Once logged into the website, state users had access to all project data for the water systems in their state and
EPA regional offices had access to the project data of states within their region. Website users were given “read
only” or “read/write” access depending on whether information posted to the website could be changed by that
entity. This created a transparent process and open communication between systems, states, and EPA, while also
maintaining a secure environment.
59
The website also served as a means of communication between survey coordinators and EPA. As EPA reviewers
completed the quality assurance reviews of each questionnaire, they uploaded the survey data to the website
database along with specific indications of any changes that had been made to the projects and why the changes
were implemented.
Each survey coordinator was able to view all projects for systems in their state and submit additional information
for projects that had been changed or deemed unallowable through EPA’s quality assurance review.
Quality Assurance
As with all earlier Assessments, the findings of the 2015 Assessment are reinforced by adherence throughout the
project to the principles embodied in the EPA’s information quality guidelines.
8
The most fundamental assurance
of the high degree of information quality is the implementation of the Agency’s Quality System. EPA implements
the system through the development of a quality assurance project plan (QAPP) for each project, which details
the specific procedures for quality assurance and quality control.
Because the Agency uses the results of this Assessment to allocate DWSRF capitalization grants to states, this
Assessment (like those that preceded it) sought to maximize the accuracy of the state-level and American Indian
and Alaska Native Village estimates of infrastructure needs. Decisions about precision levels, policies, and
procedures were established by a survey coordinators workgroup that met regularly during the 2015 Assessment.
Accuracy was maximized at the national, state, system, and project levels through the following steps. First, since
this was a sample survey, the workgroup established targets for precision of estimates in the sampling to shape
the national sample design. These precision targets are discussed in Appendix A.
Second, EPA used quality assurance procedures from the QAPP to ensure that “eligible infrastructure was clearly
defined and that documentation standards were rigorously enforced. As noted previously, for a project to be
included in the 2015 Assessment, documentation had to be submitted describing the purpose and scope of each
project. The documentation was reviewed by EPA to determine whether each project met the eligibility criteria.
The workgroup established the documentation requirements so that uniform criteria were applied to all
questionnaires.
Of the 89,728 projects submitted to the survey, EPA accepted 93 percent. The 7 percent that were not allowed
failed to meet the documentation criteria or appeared to be ineligible for DWSRF funding. Some projects were
adjusted to correct a variety of measurement problems, such as overlap between two projects (raising the issue of
double-counting), inconsistency of recorded data with project documentation, and the use of overly aggressive
(short) infrastructure life cycles by states where system planning documents were not used or available.
Third, after the survey review process, the project data were entered into a database using dual data entry
procedures to ensure correct transfer of the information. The uploaded data then went through a systematic
verification process to identify any outliers, unexpected results, or missing data. This data validation step included
8
Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility and Integrity of Information Disseminated by the
Environmental Protection Agency. https://www.epa.gov/quality/guidelines-ensuring-and-maximizing-quality-objectivity-utility-and-
integrity-information
60
queries which identify costs outliers for a given project size (high or low end); project parameter (size) outliers;
and projects where the cost provided varied substantially from what the cost model would have predicted given
the size of the infrastructure. Additional queries were run comparing total needs for a given state with other states
and with findings from previous surveys. Again, each flagged project or system was investigated and resolved.
The data were then compared at the state and national levels to identify any outliers in the data. EPA investigated
the outliers by reviewing the system’s project documentation. If the documentation did not provide enough
information to verify the project, EPA contacted the survey coordinator or the system for confirmation.
Because projects for the rehabilitation or replacement of water mains make up a high percentage of the total
national need, a high level of scrutiny was applied to these data. The submitted inventory of total pipe in a system
was reviewed using several analyses, including calculating the total amount of pipe per person and per connection
and compared to national averages, to identify changes in total amount of pipe reported for a given system between
the current and previous Assessments (for systems that participated in multiple surveys). Substantial changes in
the total amount of pipe were investigated at a state level between the current and previous surveys. With respect
to each respondent’s reported pipe rehabilitation/replacement needs, queries were run to identify the projects with
the highest reported costs, projects with the highest modeled costs, projects with the highest cost per foot of pipe,
and projects where the cost provided varied substantially from what the cost model would have predicted. Each
identified project or system was investigated and resolved, as needed. In some situations, data entry errors were
found and resolved. Often the data reviewed were determined to be valid, with the variation attributable to one or
more factors that can affect the length of pipe in a system and/or the cost of replacement, such as urban vs. rural
location or environmental conditions.
Continuing Evolution of the DWINSA
Each Assessment’s approach, policies, and guidelines have influenced the total national need and individual state
needs reported for that effort. In all cases, specific project documentation requirements and data quality objectives
were set by a workgroup including states and other stakeholders and organized by EPA. The 2003 Assessment
represented a success in better capturing long-term needs than the 1995 and 1999 efforts. The 2007 and 2011
Assessments helped guide states toward a more consistent methodology in assessing those types of needs. The
2015 Assessment maintains the methodology improvements made in 2003, 2007, and 2011 and results show an
increase in need associated with rehabilitation and replacement of existing infrastructure that may point to success
in the continued effort to better capture those long-term needs.
EPA’s quadrennial Assessment will continue to evolve, with each cycle providing valuable input as to how the
next Assessment can be improved. Each of the past Assessments and future Assessments address continuing as
well as emerging challenges for water utilities. EPA will work with the states and water utilities to improve each
survey while maintaining the integrity of the Assessment.
61
Key Observations on the Assessment’s Evolving Approach
1995 - 1999
For the first Assessment, the DWSRF was not yet in existence (1995) and for the second was in its infancy (1999). For
the 1995 Assessment, a state/EPA workgroup helped plan and design the Assessment and although some states
participated in data collection, many were unable to invest resources beyond encouraging system cooperation. For the
1999 Assessment, state programs were expected to participate in data collection. Also, by 1999, the federal DWSRF
program had been established and project-eligibility criteria were defined that specifically excluded raw water dams and
reservoirs, which had been allowed in 1995. Also at that time, the workgroup helped define Assessment policies including
those for water meters, backflow-prevention devices, and service lines. The 1999 Assessment also included needs
collected for NPNCWSs.
2003, 2007, and 2011
After the 1999 Assessment, EPA, in discussions with the water industry and the states, concluded that the DWINSA was
not fully capturing systems’ long-term needs outside of those identified in short term planning documents. This prompted
a reexamination of the survey instrument and policies to encourage systems and states to think more broadly about
existing infrastructure conditions and deficiencies. Considerable effort was invested to develop a more comprehensive
approach that included inventorying existing assets and estimating the need for rehabilitation or replacement over the
next 20 years. EPA provided flexibility to surveyed water systems and their states to forecast these longer-term needs. In
the 2003 Assessment, states and systems responded with varying approaches for identifying assets and with different
assumptions about the life cycles of those assets. These changes resulted in a significant increase in the total national
need and an increase in most states’ individual state needs. EPA’s objective to better capture the true 20-year need was
met, but the states and EPA agreed that a more consistent methodology between states should be pursued in future
Assessments.
For the 2007 Assessment, EPA and the states came to a consensus on policies to achieve consistency in both methods for
determining needs and each state’s approach to capturing those needs. Building on the methods and approaches used by
the states in the 2003 effort, EPA and the states agreed upon policies for documenting need for replacement and
rehabilitation of existing infrastructure. In 2011, the documentation requirements were further refined, incorporating three
elements: necessity, feasibility, and an indication of commitment to the project. These elements are referred to as the
weight-of-evidence determination and are further described in Appendix C.
2015
In prior surveys, EPA’s sampling method included a census of all large systems and a random sample of medium systems.
In an effort to continue building on previous efficiencies as well as to reduce burden on water systems, states, and EPA,
a modified panel approach was implemented for the statistical sampling portion of the 2015 Assessment for medium
systems. The panel (or longitudinal) approach is a statistical methodology that tracks a cross-section of observations over
time and has the added benefit of allowing for trend analyses across multiple survey cycles. With this approach, EPA
again conducted a census of large systems and reselected 75 percent of the medium systems that participated in the 2011
Assessment. Systems that were in both Assessments were asked to update their survey responses from the 2011
Assessment. This approach allowed systems to build on the efforts of the previous Assessment. EPA then selected a
random sample from among the remaining medium systems, including systems that were dropped from the 2011 sample,
to refresh the sample. This allowed some systems not selected in 2011 to be selected in 2015.
62
Appendix C Statutory and Regulatory
Criteria and Policies
The U.S. Environmental Protection Agency (EPA) recognizes that it is critical to the credibility of the 2015
Assessment, and equity between states, that EPA work with the Drinking Water Infrastructure Needs Survey and
Assessment (DWINSA or Assessment) workgroup to set clear and well-defined data collection policies, and for
EPA to apply these policies consistently to all systems. The policies are aimed at ensuring that the Assessment
meets its congressional intent, maintains the credibility of the findings, and establishes a level playing field. To
this end, the policies developed ensure two essential criteria - that only allowable needs are included, and that all
needs are adequately documented according to Assessment criteria.
Project Allowability
Because the findings of the Assessment are used to allocate Drinking Water State Revolving Fund (DWSRF)
monies, only needs associated with DWSRF-eligible projects are included in the findings. Eligibility criteria for
the DWSRF are established in the Safe Drinking Water Act (SDWA). SDWA Section 1452(a)(2) states that
DWSRF funds may be used:
“…only for expenditures (not including monitoring, operation, and maintenance
expenditures) of a type or category which the Administrator has determined, through
guidance, will facilitate compliance with national primary drinking water regulations
applicable to the system under Section 1412 or otherwise significantly further the health
protection objectives of this title....”
To be considered an allowable need, a project must be eligible for DWSRF funding, be in furtherance of the
public health protection objectives of SDWA, fall within the prescribed 20-year time frame (January 1, 2015,
through December 31, 2034), and be adequately documented.
Projects Must Be for a Capital Improvement Need
Projects that do not address a specific, tangible capital infrastructure need are not included. Non-capital needs
include operational and maintenance costs, water rights or fee payments, studies, computer software for routine
operations, and employee wages and other administrative costs.
Projects Must Be Eligible for DWSRF Funding
Projects ineligible for DWSRF funding are identified in the DWSRF regulation and include the following:
Dams or the rehabilitation of dams.
Water rights.
Raw water reservoirs or rehabilitation of reservoirs (except for finished water reservoirs and reservoirs
that are part of the treatment process and are on the property where the treatment facility is located).
Projects needed primarily for fire protection.
Projects needed primarily to serve future population growth. (Projects needed to address a deficiency
affecting current users must be sized only to accommodate a reasonable amount of population growth
expected to occur over the useful life of the facility.)
63
Projects Must Be in Furtherance of the Public Health Goals of the SDWA
Projects that are driven by objectives not based on public health protection and the goals of the SDWA are not
included in the survey. These needs can include projects for improving appearances, infrastructure demolition,
buildings and parking facilities not essential to providing safe drinking water, acquisition of land for an
unallowable project, and infrastructure needed to extend service to homes that currently have an adequate safe
drinking water supply.
Projects Must Fall Within the 20- Year Period of the Assessment
Projects for which construction began prior to January 1, 2015, and projects that are not needed until after
December 31, 2034, fell outside the time frame for the Assessment and were not included.
Projects Must Be Adequately Documented
Project documentation is a critical piece of the Assessment’s credibility and fairness. It is described in more detail
later in this Appendix.
Other Unallowable Needs
Besides the project criteria discussed above, other limitations established by the workgroup were:
Infrastructure needs that occur more than once during the 20-year survey period could be listed only once
on the survey.
Multiple projects meeting the same need, such as rehabilitating a tank and later replacing the same tank,
could not all be included.
Projects driven solely by a non-water-related issue such as highway relocation were not included.
Projects to acquire existing infrastructure were not considered capital infrastructure costs.
Most vehicles and tools were considered operation and maintenance costs.
Projects that are not the responsibility of the public water system.
If projects associated with an unallowable need were submitted, they were excluded from the Assessment by
EPA. EPA understands that these projects often represent legitimate and even critical needs that a water system
must pursue to continue to provide service to its customers. However, because they do not meet the allowability
criteria they are not the subject of the DWINSA.
Documentation Requirements
The 2015 Assessment essentially maintained the documentation requirements established for the 2003
Assessment and improved upon by the 2007 and 2011 Assessment efforts. In particular, EPA and the workgroup
came to consensus to incorporate the same improvements used by the 2007 and 2011 Assessments to ensure a
consistent approach to data collection and to the assessment of need applied by each survey coordinator.
High-quality documentation is required to justify the need for a project, defend cost estimates provided by the
water system, provide a defensible assessment of national need, and ensure fair allotment of DWSRF monies.
The documentation of need and cost for each project was carefully reviewed to ensure that the criteria set in the
Assessment approach and established by consensus of EPA and the workgroup were met.
64
Types of Documentation
In an effort to ensure more consistency in each state’s approach to the assessment of its water systems’ needs, the
workgroup defined for the 2007 Assessment, and retained for the 2011 and 2015 efforts, three types of
documentation that could be provided to describe a need or provide a cost:
Independent Documentation. A document or report generated through a process independent of the
Assessment. Because these documents were not generated specifically for the Assessment, it is assumed
that there is no intentional bias of over reporting of need.
Survey-generated Documentation. A statement or document discussing the need for a project generated
specifically for the Assessment by the system or the state.
Combination Documentation. A combination of independent and survey-generated documentation to
justify project need or cost. Independent documentation does not always directly address the reason a
project is being pursued by a system and therefore may not fully establish that the project meets the
survey’s allowability criteria. Systems often added survey-generated documentation to independent
documents to clarify the need for the project.
Documentation of Need
Documentation explains the scope of the project, explains why the project is needed, and gives an indication of
the public health need that would be addressed by the project. In order for the project to be accepted, the
documentation of need must:
Provide sufficient information for EPA to review the allowability of the project.
Provide adequate data to check the accuracy of the data entered on the questionnaire.
Be dated and be less than four years old.
The type of documentation required varied by the specific project type. Minimum requirements were set to allow
a minor level of effort by states and water systems to document straight-forward projects. Doing so made more
resources available to identify and document projects for which allowability was more questionable. Projects fell
into the following levels of documentation requirements:
Projects that required independent documentation of need.
Projects for which survey-generated documentation were permitted but to which a weight of evidence
review was applied.
Projects accepted with any forms of documentation.
The level of documentation required depended on the type of project and whether the project was for new
infrastructure or for the replacement, rehabilitation, or expansion/upgrade of existing infrastructure. Any of the
three forms of documentation were acceptable for projects to rehabilitate or replace infrastructure assumed to
have a life-cycle of 20 years or less.
65
Projects likely to be driven by a need that is not DWSRF-eligible (such as to accommodate growth or meet fire
suppression needs) generally require independent documentation. Most projects for the installation of new
infrastructure fall into this category. For those projects, such as the construction of a new treatment system or new
storage tank, EPA reviews independent documentation and applies a “weight-of-evidence” approach to determine
whether the project could be included in the Assessment.
Projects for Which Independent
Documentation Is Required
Generally, projects that could be considered
unallowable (such as projects to meet anticipated
growth) or that are for infrastructure likely to have
an expected life of more than 20 years (such as a
water main) require independent documentation of
need. EPA and the workgroup assumed that systems
pursuing needs in this category are often in the
process of formal planning and therefore
independent documents are likely to exist. Projects
requiring independent documentation for the 2015
Assessment included the following:
Sources installation of new surface water
intakes, off-stream raw water storage, or
new aquifer storage and recovery wells.
Treatment installation, replacement, or
expansion/upgrade of a complete treatment
plant or new treatment components.
Storage installation of new elevated or
ground-level finished water or treated water
storage.
Pipe installation of new water mains,
rehabilitation, and replacement of a
substantial portion (in excess of 10 percent of the total) of the system’s existing water mains.
Pumping installation of new pump stations.
Weight of Evidence
Documentation must include adequate system-specific and
project-specific details to verify that the project meets the
allowability criteria and that the project is needed. For the
2011 Assessment, three specific weight of evidence criteria
had to be supported by documentation. The project had to
be shown to be:
--Necessary to meet the requirements of the SDWA and for
public health purposes;
--Feasible by being typical of today’s water engineering
standards and practices; and
--Committed to by relevant decision-makers as specified in
supporting documents or by a standing history of such
commitment to similar projects, as common practice by the
industry, or made evident in the documentation as a
standing policy by the specific water system, state, or other
relevant authority.
The 2015 Assessment continued the practice used since the
2003 effort to better capture 20-year investment needs by
including an asset inventory-based approach to identify
long-term infrastructure replacement and rehabilitation
needs (see Appendix B).
66
Projects for Which Survey-Generated Documentation Is Allowed, but a Weight of Evidence Review Is
Applied
Needs that are subject to a weight-of-evidence review include projects that are significant in scope or that could
be for unallowable need (such as anticipated growth), but are not necessarily likely to be included in a planning
document. For these projects, systems are asked to provide enough information for the reviewer to ascertain
whether the project is for an allowable need. These projects include the following:
Sources construction of new wells or springs, new well pumps or raw water pumps, and replacement
or rehabilitation of any source.
Storage replacement of a finished water elevated or ground level storage tank or installation of a new
hydro-pneumatic storage tank.
Pipe a significant amount of new water main appurtenances such as valves, or backflow prevention
devices, or replacement of over 10 percent of the existing inventory of those items.
Pumping replacement of an existing pump station or installation of a new finished water pump.
Emergency Power new emergency power generators.
Projects for Which A ll Forms of Documentation Are Accepted
Projects for infrastructure that is generally expected to require rehabilitation or replacement within a 20-year
period are accepted with minimum documentation of need. Survey-generated documentation, including a water
system’s inventory of existing infrastructure assets, was sufficient for these projects, which include:
Sources replacement or rehabilitation of well pumps, raw water pumps, and other miscellaneous
source projects.
Treatment rehabilitation of a complete treatment plant, or rehabilitation or replacement of treatment
components, or replacement of treatment monitors.
Storage rehabilitation of any finished water storage tank or cistern, cover of finished water storage
tank, replacement of hydro pneumatic tanks.
Pumping replacement or rehabilitation of any pump, or rehabilitation of any pump station.
Pipe rehabilitation or replacement of water mains up to 10 percent of the system’s existing total pipe
inventory.
Other infrastructure such as replacement of lead service lines and installation of control valves,
backflow prevention, meters, controls, and replacement of emergency power.
Documentation of Cost
To estimate a 20-year national, American Indian, Alaska Native, and individual state need, every project must
have an estimated cost. There were two primary methods for assigning costs to a project:
Systems provided an independent cost estimate.
Systems provided adequate information for EPA to estimate a cost using a cost model.
67
For systems that provide a cost estimate, the documentation must:
Include the date the estimate was derived.
Be generated through a process independent of the Assessment.
Be no more than 10 years old (earlier than January 1, 2005).
Not include loan origination fees, finance charges, bond issuance fees or costs, interest payments on a
loan, or inflationary multipliers for future projects.
Since projects with adequately documented costs are the basis of the cost models, systems are encouraged to
provide both cost and design parameters for as many projects as possible so that the data can be used to update
existing cost models.
If a cost is not provided, key information on design parameters and project type is required for EPA to assign a
cost to the project using a cost model. However, EPA is unable to model a few types of infrastructure projects
(i.e., projects that were too unique or site-specific). In those cases, a documented cost estimate was required in
order for the cost to be included in the Assessment.
68
Appendix D - Accuracy, Precision, and
Uncertainty
Uncertainty, precision, and bias affect the accuracy of an estimate based on a statistical sample. While a sample
can be designed to meet certain precision targets, other sources of uncertainty and potential biases may diminish
the accuracy of estimates.
Uncertainty
There are two types of uncertainty at play in the Drinking Water Infrastructure Needs Survey and Assessment
(DWINSA or Assessment). Real uncertainties are created as survey respondents predict future needs. The U.S.
Environmental Protection Agency (EPA) is asking systems not only to provide their existing needs, but also to
anticipate what their future needs will be. It is difficult to predict future needs. Since no one knows, for example,
when a pump will fail or exactly what it will cost to fix or replace it when it does fail, there is real uncertainty
about the accuracy of estimates of future investment needs.
A second source of uncertainty is the use of a probability sample to estimate need. Uncertainties are created due
to the inherent limitations of statistical analyses. The use of a random sample and cost models create such
stochastic (i.e., random or arising from chance) uncertainties in the survey. In assessing the impact that the sample
has on the estimate, EPA distinguishes between two sources of stochastic uncertainty: precision and bias.
Precision
Precision is the degree to which additional measurements would produce the same or similar results. Two factors
affect the precision of sample-based estimates. First is the inherent variability of the data. If systems’ needs are
similar, the margin of error will be smaller than if needs vary greatly across systems. The second factor is the size
of the sample. Larger samples produce more precise estimates than smaller ones.
The use of a random sample introduces uncertainty to the estimate. A different sample would lead to a different
estimate of each state’s need, since there will always be some variability among different systems selected in a
sample. Because the Assessment relies on a random sample, the sample should provide an unbiased estimate of
the total need. The level of confidence in the estimate is reflected in the confidence interval.
EPA’s goal is to be 95 percent confident that the margin of error for the survey is ± 10 percent of the total need
for systems serving more than 3,300 people for each fully surveyed state and for all American Indian and Alaska
Native Village public water systems, assuming that the data provided are unbiased. (The estimates for individual
partial participation states do not meet these precision targets. The Assessment also has separate precision targets
for systems in the state survey serving 3,300 or fewer people.)
If the systems that responded to the survey reported the cost of their investment needs for all projects, sampling
error would be the only stochastic source of uncertainty. But systems do not have cost estimates for most of the
projects they reported. EPA imputed the cost of these projects using cost models based on cost estimates submitted
for other projects. As with sampling, there is a degree of predictable error associated with such modeling.
69
Bias
Sampling error is random. It is as likely to lead to an estimate that is greater than the true value as it is lower than
the true value. Bias, however, is not random. An estimator is biased if its expected value is different from the true
value. An estimator is upwardly biased if it consistently leads to an estimate that is greater than the true value. It
is downwardly biased if it consistently leads to an estimate that is less than the true value. The Assessment has
both upward and downward biases. EPA implemented policies and procedures to mitigate the impact of these
biases.
Downward bias
Past Assessments and studies of these Assessments have shown that systems are likely to underestimate their
needs. There is little theory or empirical evidence to suggest that systems overstate their needs. This
understatement is brought on for two primary reasons. One is that the bulk of a system’s infrastructure is
underground in the form of transmission and distribution mains. It is difficult to assess the need for addressing
these out-of-sight assets. The second is that the survey assesses systems’ 20-year need. Many systems have not
undertaken the long-term planning necessary to identify future infrastructure needs.
Upward bias
In part to help address the downward bias introduced by systems’ underestimating their needs, EPA enlisted the
help of states in the data collection effort. However, because these entities are the recipients of the capitalization
grants determined by the Assessment, there is an incentive for them to overestimate their systems’ needs. This
situation introduces a possible upward bias in the estimate of the needs generated by systems with this type of
input.
It is unlikely that this bias applies to the Assessment estimate of small system need in the state survey. The small
system survey is conducted by EPA, without states’ direct involvement. For this reason, there is no upward bias
in this portion of the survey. In addition, because these small system surveys are conducted by trained
professionals, EPA expects very little downward bias.
Twenty-two states, the U.S. territories, and the District of Columbia have needs of less than 1 percent of the
national need. These states receive the minimum Drinking Water State Revolving Fund (DWSRF) allocation
regardless of the need reported (1 percent for states, Puerto Rico, and the District of Columbia; 1.5 percent for
U.S.
territories). For this reason, there is likely no upward bias in the allocation for these states, and only the
downward bias discussed above influences need in these states.
With input from states as well as a peer-review process for the 2007 Assessment, EPA implemented policies to
help address both upward and downward bias. These policies included:
Projects to rehabilitate or replace infrastructure generally considered in need of attention within a 20-year
period were allowed based on system- or other entity-signed statements and project descriptions. Systems
were encouraged to consider their entire inventory and document all such needs if legitimate.
Projects to rehabilitate or replace infrastructure not necessarily considered in need of attention within a
20-year period were allowed with documentation independent of the Assessment or a system or other
entity’s statement if the documentation included additional project-specific information such as an
assessment of age, current condition, and maintenance history.
Projects that include the installation/construction of new infrastructure generally received a high degree
of scrutiny to ensure that they met allowability criteria.
70
Some infrastructure was only allowed if independent documentation was provided. This included new
surface water sources, new treatment plants or components, the replacement or expansion of an existing
treatment plant, new storage tanks, and widespread replacement or rehabilitation of the distribution system
(defined as more than 10 percent of the existing pipe inventory).
71
Appendix E - Summary of Findings for
State Systems Serving 10,000 and
Fewer People
Community Water Systems Serving 10,000 and Fewer People
The Safe Drinking Water Act requires that states use at least 15 percent of their Drinking Water State Revolving
Fund funding for financial assistance to community water systems (CWSs) serving populations of 10,000 and
fewer. Of the $459.4 billion in need for all CWS in states, those serving 10,000 and fewer people represent 29
percent or approximately $132.3 billion of needs (this number includes CWSs in U.S. territories). Exhibit E.1
presents the 20-year needs for these smaller community systems by state and project category. It also compares
the reported need of these systems to the state’s total CWS need. All data in Exhibit E.1 exclude needs related to
not-for-profit noncommunity, American Indian, and Alaska Native Village water systems.
Exhibit E.1: State Need by Project Category for CWSs Serving a Population of 10,000 and Fewer
Compared to All CWSs (in millions of 2015 dollars)
State
CWSs Serving 10,000 or Fewer People
Total 20-
Year Need
of All
CWS*
% of CWS
Need
Related to
Systems
Serving
10,000 or
Fewer
People*
Transmission
and
Distribution
Source
Treatment
Storage
Other
Total 20-
Year Need of
CWS
Serving
10,000 or
Fewer
People*
Alabama
$1,891.6
$45.1
$145.3
$196.6
$39.0
$2,317.6
$11,257.0
20.6%
Arizona
$1,047.9
$114.7
$214.4
$256.8
$16.1
$1,649.9
$9,107.1
18.1%
Arkansas
$2,062.2
$121.5
$317.4
$333.9
$52.9
$2,887.8
$7,367.6
39.2%
California
$3,613.0
$539.8
$1,393.5
$1,170.8
$69.3
$6,786.3
$50,904.9
13.3%
Colorado
$2,060.8
$144.8
$674.8
$363.0
$38.7
$3,282.1
$10,187.0
32.2%
Connecticut
$571.0
$87.0
$151.7
$139.0
$22.1
$970.8
$3,982.9
24.4%
District of
Columbia
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$1,741.9
0.0%
Florida
$1,773.6
$290.6
$435.0
$367.3
$62.8
$2,929.3
$21,724.8
13.5%
Georgia
$2,103.2
$244.9
$414.7
$462.8
$42.6
$3,268.1
$12,442.7
26.3%
Illinois
$4,641.4
$339.0
$853.7
$811.5
$86.2
$6,731.8
$20,771.0
32.4%
Indiana
$1,676.3
$155.3
$278.0
$305.4
$30.1
$2,445.1
$7,297.0
33.5%
Iowa
$2,514.1
$232.0
$425.9
$377.2
$32.1
$3,581.2
$7,829.1
45.7%
Kansas
$2,063.1
$164.4
$363.3
$323.0
$28.4
$2,942.2
$5,318.2
55.3%
Kentucky
$1,342.1
$46.8
$141.3
$178.9
$16.4
$1,725.5
$8,230.7
21.0%
Louisiana
$2,374.6
$212.8
$466.3
$408.7
$70.8
$3,533.2
$7,311.7
48.3%
Maine
$456.7
$59.1
$124.7
$118.5
$12.1
$771.1
$1,304.7
59.1%
72
State
CWSs Serving 10,000 or Fewer People
Total 20-
Year Need
of All
CWS*
% of CWS
Need
Related to
Systems
Serving
10,000 or
Fewer
People*
Transmission
and
Distribution
Source
Treatment
Storage
Other
Total 20-
Year Need of
CWS
Serving
10,000 or
Fewer
People*
Maryland
$575.7
$85.3
$150.6
$141.6
$15.1
$968.3
$9,205.8
10.5%
Massachusetts
$1,061.0
$109.7
$333.7
$288.8
$58.0
$1,851.3
$12,202.8
15.2%
Michigan
$2,518.6
$324.9
$494.1
$447.5
$54.0
$3,839.1
$12,448.6
30.8%
Minnesota
$1,687.4
$225.4
$412.3
$372.0
$56.9
$2,753.9
$7,168.4
38.4%
Mississippi
$2,216.4
$249.9
$489.6
$438.4
$38.0
$3,432.2
$4,811.1
71.3%
Missouri
$2,769.3
$279.6
$573.9
$545.5
$27.8
$4,196.2
$8,869.3
47.3%
Nevada
$372.7
$49.3
$134.4
$114.4
$7.6
$678.3
$5,298.7
12.8%
New Jersey
$848.9
$84.4
$142.8
$210.7
$20.6
$1,307.4
$8,325.5
15.7%
New York
$3,496.6
$433.3
$994.6
$793.5
$69.2
$5,787.1
$22,606.3
25.6%
North
Carolina
$2,491.4
$392.5
$641.0
$496.6
$85.2
$4,106.7
$16,254.3
25.3%
Ohio
$2,019.4
$250.8
$605.5
$454.7
$53.7
$3,384.1
$13,048.3
25.9%
Oklahoma
$2,577.1
$187.8
$459.3
$421.7
$46.5
$3,692.5
$6,830.7
54.1%
Oregon
$1,173.0
$164.4
$394.7
$343.4
$39.6
$2,115.2
$6,180.0
34.2%
Pennsylvania
$3,232.4
$380.2
$1,011.0
$805.2
$139.0
$5,567.8
$16,415.0
33.9%
Puerto Rico
$558.3
$73.2
$221.7
$147.0
$13.8
$1,014.0
$3,702.6
27.4%
South
Carolina
$1,192.3
$93.0
$225.6
$171.9
$30.7
$1,713.6
$6,106.9
28.1%
Tennessee
$1,552.4
$60.0
$192.8
$185.3
$5.0
$1,995.5
$8,727.3
22.9%
Texas
$8,450.2
$793.0
$1,852.4
$1,905.0
$268.5
$13,269.0
$45,091.0
29.4%
Utah
$686.3
$109.8
$189.2
$213.1
$19.5
$1,217.9
$4,338.0
28.1%
Virginia
$1,603.1
$171.9
$499.3
$421.5
$64.9
$2,760.6
$8,019.1
34.4%
Washington
$2,549.6
$418.4
$677.9
$653.6
$73.5
$4,373.0
$11,582.4
37.8%
Wisconsin
$1,447.0
$213.2
$509.7
$444.5
$18.5
$2,632.9
$7,957.4
33.1%
Partial
Participation
States
$8,431.1
$1,013.8
$2,097.0
$1,834.2
$208.0
$13,584.1
$26,437.4
51.4%
Subtotal
$83,701.4
$8,961.4
$19,703.2
$17,663.5
$2,033.1
$132,062.5
$458,405.6
28.8%
American
Samoa
$29.6
$3.8
$10.8
$7.5
$0.5
$52.2
$302.5
17.3%
Guam
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$270.0
0.0%
North
Mariana Is.
$49.6
$7.5
$14.4
$12.5
$1.1
$85.1
$198.4
42.9%
Virgin Islands
$41.2
$8.4
$11.8
$9.6
$1.2
$72.2
$235.1
30.7%
Subtotal
$120.4
$19.6
$37.0
$29.6
$2.8
$209.4
$1,005.9
20.8%
Total
$83,821.8
$8,981.0
$19,740.2
$17,693.1
$2,035.9
$132,271.9
$459,411.4
28.8%
*
Excludes NPNCWS
The needs for states that opted out of the medium portion of the survey are presented cumulatively and not by state. The list of partial
participation states is shown in Exhibit 2.4.
73
Glossary
Capital Improvement Plan (CIP): a document produced by a local government, utility, or water system that
thoroughly outlines, for a specified period of time, all needed capital projects, the reason for each project, and the
projects’ costs.
Coliform bacteria: a group of bacteria whose presence in a water sample indicates the water may contain disease-
causing organisms.
Community water system (CWS): a public water system that serves at least 15 connections used by year-round
residents or that regularly serves at least 25 residents year-round. Examples include cities, towns, and
communities such as retirement homes.
Current infrastructure needs: new facilities or projects to address deficiencies in existing facilities for which
water systems would begin construction as soon as possible to avoid a threat to public health.
Engineer’s report: a document produced by a professional engineer that outlines the need and cost for a specific
infrastructure project.
Existing infrastructure: Infrastructure that was in place, fully installed and providing service to the water utility
prior to the commencement of this survey.
Existing regulations: drinking water regulations promulgated by EPA under the authority of the Safe Drinking
Water Act; existing regulations can be found at Title 40 Part 141, the Code of Federal Regulations (40 CFR 141).
Finished water: water that is considered safe to drink and suitable for delivery to customers.
Future infrastructure needs: infrastructure deficiencies that a system expects to address in the next 20 years
because of predictable deterioration of facilities. Future infrastructure needs do not include current infrastructure
needs. Examples are storage facility and treatment plant replacement, where the facility currently performs
adequately but will reach the end of its useful life in the next 20 years. Needs solely to accommodate future growth
are not included in the Assessment.
Ground water: any water obtained from a source beneath the surface of the ground, which has not been classified
as ground water under the direct influence of surface water.
Growth: The expansion of a water system to accommodate or entice future additional service connections or
consumers. Needs planned solely to accommodate projected future growth are not included in the Assessment.
Eligible projects, however, can be designed for growth expected during the design-life of the project. For example,
the Assessment would allow a treatment plant needed now and expected to treat water for 20 years. Such a plant
could be designed for the population anticipated to be served at the end of the 20-year period.
Infrastructure needs: the capital costs associated with ensuring the continued protection of public health through
rehabilitation or construction of facilities needed for continued provision of safe drinking water. Categories of
infrastructure need include source development and rehabilitation, treatment, storage, and transmission and
distribution. Operation and maintenance needs are not considered infrastructure needs and are not included in this
document.
74
Large water system: in this document, this category comprises community water systems serving more than
100,000 people.
Medium water system: in this document, this category comprises community water systems serving from 3,301
to 100,000 people.
Microbiological contamination: the occurrence of protozoan, bacteriological, or viral contaminants in a water
supply.
New infrastructure: Infrastructure that was not in place and was not providing service to the water utility prior
to the commencement of this survey.
Noncommunity water system: a public water system that is not a community water system and that serves a
nonresidential population of at least 25 individuals daily for at least 60 days of the year. Examples of not-for-
profit noncommunity water systems include schools and churches.
Public water system: a system that provides water to the public for human consumption through pipes or other
constructed conveyances, if such system has at least 15 service connections or regularly serves an average of at
least 25 individuals daily at least 60 days out of the year.
Regulatory need: a capital expenditure required for compliance with Safe Drinking Water Act regulations.
Safe Drinking Water Act (SDWA): a law passed by Congress in 1974 and amended in 1986 and 1996 to ensure
that public water systems provide safe drinking water to consumers (42 U.S.C.A. §300f to 300j-26).
Small water system: in this document, this category comprises community water systems serving up to 3,300
people.
Source: a project category that includes the costs associated with developing or improving sources of water for
public water systems.
State: in this document, state refers to all 50 states of the United States plus Puerto Rico, the District of Columbia,
American Samoa, Guam, the Commonwealth of Northern Mariana Islands, and the U.S. Virgin Islands.
Storage: a project category that addresses finished water storage for public water systems.
Supervisory Control and Data Acquisition (SCADA): an advanced control system that collects all system
information and allows an operator, through user-friendly interfaces, to view all aspects of the system from one
place.
Surface water: all water that is open to the atmosphere and subject to surface run-off, including streams, rivers,
and lakes.
Transmission and distribution: a project category that includes installation, replacement, or rehabilitation of
transmission or distribution mains and associated appurtenances that carry drinking water from the source to the
treatment plant or from the treatment plant to the consumer, as well as pump stations in the distribution system.
Treatment: a project category that includes conditioning water or removing microbiological or chemical
contaminants. Filtration of surface water, pH adjustment, softening, and disinfection are examples of treatment.
75
Ductile iron pipe for a raw water line to a new water treatment
plant in Arkansas City, KS.
76